Ontolog Forum - User contributions [en]

ConferenceCall 2025 03 26

2025-03-25T05:12:56Z

AndreaWesterinen: /* Agenda */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Track 3]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::26 Mar 2025 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2025|Ontology Summit 2025]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''[[AndreaWesterinen|Andrea Westerinen]]''' - ''Narratives to Gain Situational Awareness via Ontologies''
* Abstract: This presentation introduces an ontology to describe how we conceptualize and communicate about situations in the real-world - the events, entities and relationships that exist. Building on ontological semantics, we propose a mapping from narratives (descriptions of the real-world) to propositional statements (triples), where the truth values of the propositions depend on objective reality but filtered by the subjective and intersubjective interpretations and goals of the narrator. Inspired by Walter Fisher’s narrative paradigm, we highlight the universality of storytelling as a means for humans to organize, interpret, and communicate experience. To model and compare narrative structures, we present an ontology that captures the who, what, where, when, why, and how within narratives. This enables systematic exploration of narratives' sources, sentiments, use of language and much more. And, this exploration can occur across multiple narratives and across time. Using an ontology provides conceptual clarity and minimizes ambiguity. The approach enables comparison of the topics, entities and actions discussed and how the narrative is structured and conveyed in order to appeal and engage the reader/listener.

== Conference Call Information ==
* Date: '''Wednesday, 26 Mar 2025'''
* Start Time: 9:00am PDT / 12:00pm EDT / 5:00pm CET / 4:00pm GMT / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=3&day=26&year=2025&hour=12&min=00&sec=0&p1=179 World Clock]
** Note: The US and Canada are on Daylight Saving Time while Europe has not yet changed.
* Expected Call Duration: 1 hour
{{:OntologySummit2025/ConferenceCallInformation}}

== Participants ==

== Discussion ==

== Resources ==

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2025]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2025_03_25]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2025]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2025_03_27]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2025]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

OntologySummit2025

2025-03-25T05:08:53Z

AndreaWesterinen: Added Track 3 info for 26 March

= Ontology Summit 2025 =
== Conceptualization, Analysis and Formalization ==
=== The Two Sides of Ontology: Relating ontologies to the world and to theories about the world ===

The [[OntologySummit|Ontology Summit]] is an annual series of events that involves the ontology community and communities related to each year's theme chosen for the summit. The Ontology Summit was started by Ontolog and NIST, and the program has been co-organized by Ontolog and NIST along with the co-sponsorship of other organizations that are supportive of the Summit goals and objectives.

As part of Ontolog’s general advocacy to bring ontology science and related engineering into the mainstream, we endeavor to facilitate discussion and knowledge sharing amongst stakeholders and interested parties relevant to the use of ontologies. The results will be synthesized and summarized in the form of the Ontology Summit 2024 Communiqué, with expanded supporting material provided on the web and in journal articles.

= Process and Deliverables =
Similar to our last 19 summits, this [[OntologySummit2025|Ontology Summit 2025]] will consist of virtual discourse (over our archived mailing lists), virtual presentations and panel sessions as part of recorded video conference calls.
As in prior years the intent is to provide some synthesis of ideas and draft a communiqué summarizing major points.

Meetings are at Noon US/Canada Eastern Time on Wednesdays and last about an hour.
The sessions are [https://us02web.zoom.us/j/88593616861?pwd=HafnK0yB7PFDK1EyiUyQRDKanZlbjU.1 Zoom Meetings].

== Description ==

In this summit we will consider the question of what an ontology is as well as how ontologies are related to other notions such as conceptualizations, theories and semantics.

For an overview of Track 1 "Introduction to the Summit and Track 1" by Gary Berg-Cross see https://ontologforum.s3.us-east-1.amazonaws.com/OntologySummit2025/Launch/Track-1-Launch--GaryBerg-Cross_20250115.pdf

'''[[NicolaGuarino|Nicola Guarino]]''' will set the stage for the summit with his keynote address: "Ontologies as specifications of conceptualizations: correctness, precision, and accuracy", which
will be elaborated by '''[[GiancarloGuizzardi|Giancarlo Guizzardi]]''' who will discuss semantics, ontology and explanation.
Accordingly, conceptualization is fundamental for ontologies, but a careful analysis is necessary for a specification to be useful.
'''[[MichaelGruninger|Michael Gruninger]]''' and '''[[BarrySmith|Barry Smith]]''' will then examine how one can specify the conceptualization of reality by means of mathematical theories and axioms as well as the limits of such approaches.
The next session will raise the question of what a theory is, which will
segue to a series of sessions that survey general philosophical and theoretical issues.

The second half of the summit will survey more concrete issues, specifically about data and its relationship to conceptualizations, reality and ontologies.
Of special interest are ontologies that have large amounts of continually increasing instance data.
How can one effectively verbalize and visualize such large ontologies?
How can one control the quality as the data expands?
How effective are these ontologies in practice?
Can the ontologies adequately support reasoning?

== Organization ==

We will begin with a [[ConferenceCall_2025_01_15|Overview Session]] on Wednesday, 15 January 2025.
This will be followed by a [[ConferenceCall_2025_01_22|Keynote Address "Ontologies as specifications of conceptualizations: correctness, precision, and accuracy”.]] featuring '''[[NicolaGuarino|Nicola Guarino]]''' on Wednesday, 22 January 2025.
The summit will consist of four tracks as follows:

=== Track 1: Conceptualizing the theoretical form of reality ===
Track Chair: [[GaryBergCross|Gary Berg-Cross]]

* [[ConferenceCall_2025_01_29|29 January 2025]] '''[[GiancarloGuizzardi|Giancarlo Guizzardi]]'''
** Title: "Explanation, Semantics, and Ontology"
** Abstract: It is well-known by now that, of the so-called 4Vs of Big Data (Velocity, Volume, Variety and Veracity), the bulk of effort and challenge is in the latter two: (1) data comes in a large variety of representations (both from a syntactic and semantic point of view); (2) data can only be useful if truthful to the part of reality that it is supposed to represent. Moreover, the most relevant questions we need to have answered in science, government and organizations can only be answered if we put together data that reside in different data silos, which are produced in a concurrent manner by different agents and in different points of time and space. Thus, data is only useful in practice if it can (semantically) interoperate with other data. Every data schema represents a certain conceptualization, i.e., it makes an ontological commitment to a certain worldview. Issue (2) is about understanding the relation between data schemas and their underlying conceptualizations. Issue (1) is about safely connecting these different conceptualisations represented in different schemas. To address (1) and (2), we need to be able to properly explain these data schemas, i.e., to reveal the real-world semantics (or the ontological commitments) behind them. In this talk, I discuss the strong relation between the notions of real-world semantics, ontology, and explanation. I will present a notion of explanation termed Ontological Unpacking, which aims at explaining symbolic representation artifacts (conceptual models connected to data schemas, knowledge graphs, logical specifications). I show that these artifacts when produced by Ontological Unpacking differ from their traditional counterparts not only in their expressivity but also on their nature: while the latter typically merely have a descriptive nature, the former have an explanatory one. Moreover, I show that it is exactly this explanatory nature that is required for semantic interoperability. I will also discuss the relation between Ontological Unpacking and other forms of explanation in philosophy and science, as well as in Artificial Intelligence. I will argue that the current trend in XAI (Explainable AI) in which “to explain is to produce a symbolic artifact” (e.g., a decision tree or a counterfactual description) is an incomplete project resting on a false assumption, that these artifacts are not “inherently interpretable”, and that they should be taken as the beginning of the road to explanation, not the end. This talk is based on the following paper: https://www.sciencedirect.com/science/article/pii/S0169023X24000491

* [[ConferenceCall_2025_02_05|5 February 2025]] '''[[MichaelGruninger|Michael Gruninger]]'''
** Title: The Heirs of Hilbert's Sixth Problem
** Abstract: In an address to the International Congress on Mathematicians in 1900, David Hilbert posed twenty-three challenge problems, in areas ranging from logic to number theory and partial differential equations. These problems have had a profound impact on research in mathematics. However, the sixth problem posed by Hilbert has never been adequately addressed: ``Mathematical treatment of the axioms of physics: The investigations on the foundations of geometry suggest the problem: To treat in the same manner, by means of axioms, those physical sciences in which mathematics plays an important part." This talk will explore the ways in which ontologies are the axiomatic theories required by Hilbert as a solution to his Sixth Problem. It will also consider how the methodology for evaluating scientific theories can be applied to the problem of empirical evaluation of ontologies.
* [[ConferenceCall_2025_02_12|12 February 2025]] '''[[BarrySmith|Barry Smith]]'''
** Title: Models, theories and ontologies
** In the paper https://arxiv.org/abs/2305.01560 Jobst Landgrebe and I outline the beginnings of on ontology of physics and mathematics from a BFO (= commonsensical) perspective. I will sketch how the ontologies of classical and modern physics relate to the ontology of common sense and of mathematics. In brief, classical physics inherits the common-sense view of nature, and uses mathematics to formalise our natural understanding of the causes and effects we observe in time and space when we select subsystems of nature for modelling. But in modern physics, we do not extend the realm of common sense by augmenting our knowledge of what is going on in nature. Rather, we have measurements that we do not understand, so we know nothing about the ontology of what we measure. We help ourselves by using entities from mathematics, which we do understand ontologically.
* [[ConferenceCall_2025_02_19|19 February 2025]] '''[[KenBaclawski|Ken Baclawski]]'''
** Title: Conceptualizing Domains
** Abstract: The summit so far has presented some very important topics for our community. This talk will summarize and expand on some of what has transpired in the sessions as well as the forum discussions. The issues that I will present include definitions of theory, observer effects, unreasonable effectiveness, semantics of voids and the Cantor paradise. Many of these issues represent new challenges and opportunities for ontologies. I will end with an introduction to the upcoming tracks.

=== Track 2: Theoretical Knowledge and Reality ===
Track CoChairs: [[KennethBaclawski|Ken Baclawski]] and [[AlexShkotin|Alex Shkotin]]

* [[ConferenceCall_2025_02_26|26 February 2025]] '''[[ArunMajumdar|Arun Majumdar]]''' and '''[[JohnSowa|John Sowa]]''' - ''Generating Ontologies''
* Abstract: The LLMs of Generative AI support powerful methods for translating, finding, combining, and transforming data represented in languages of any kind — natural, artificial, linear, diagrammatic, pictorial, or structural in any number of dimensions. Their reasoning is based on abduction or educated guessing. But abduction alone cannot evaluate the accuracy of the guesses. Therefore, many generative AI systems are hybrids that also use traditional symbolic AI methods for detecting and avoiding errors. Permion technology is based on an integrated combination of the best generative and symbolic methods. Reasoning is performed in a cognitive cycle of abduction, deduction, testing, induction, and repeat. The cognitive cycle begins with a basic ontology, which it revises and extends dynamically.

* [[ConferenceCall_2025_03_05|5 March 2025]] '''Øystein Linnebo''' ''Constructional ontology and criteria of identity''
* Abstract: Gödel and others have suggested that a set can be regarded as constructed from its elements. Given any objects, we can apply the “set of” operation to construct the set of these objects. Inspired by Gödel’s suggestion, this talk presents a far more general constructional approach to ontology. First, I clarify what it is for some objects to be “constructed” from others. The key is that all truths about the “new” objects that are constructed must reduce to truths about the “old” objects on which the construction is based. Then, I explain the central role of criteria of identity in the constructional approach. When constructing “new” objects, it is particularly important to stipulate what it takes for them to be identical or distinct. A variety of examples are provided, including mereological sums, ordered pairs, as well as cardinal and ordinal numbers. Finally, I present a logical framework in which this constructional approach can be developed. In this framework, a large and natural family of forms of construction can be proved to be consistent.

* [[ConferenceCall_2025_03_12|12 March 2025]] '''Kit Fine'''

=== Track 3: From Reality to Data ===
* Track Chair: Mike Bennett

* [[ConferenceCall_2025_03_26|26 March 2025]] '''[[AndreaWesterinen|Andrea Westerinen]]'''
** Title: Narratives to Gain Situational Awareness via Ontologies
** Abstract: This presentation introduces an ontology to describe how we conceptualize and communicate about situations in the real-world - the events, entities and relationships that exist. Building on ontological semantics, we propose a mapping from narratives (descriptions of the real-world) to propositional statements (triples), where the truth values of the propositions depend on objective reality but filtered by the subjective and intersubjective interpretations and goals of the narrator. Inspired by Walter Fisher’s narrative paradigm, we highlight the universality of storytelling as a means for humans to organize, interpret, and communicate experience. To model and compare narrative structures, we present an ontology that captures the who, what, where, when, why, and how within narratives. This enables systematic exploration of narratives' sources, sentiments, use of language and much more. And, this exploration can occur across multiple narratives and across time. Using an ontology provides conceptual clarity and minimizes ambiguity. The approach enables comparison of the topics, entities and actions discussed and how the narrative is structured and conveyed in order to appeal and engage the reader/listener.

This track will cover a range of topics and questions such as:
* Truthmakers; perception; situation awareness
* Quality control of ontologies from the point of view of supporting theories
* How can we make our devices and manipulators?
* How can we use theoretical knowledge to create our measurement and other tools?
* Data verbalization: any unit of data can be read out loud.
* Data visualization

=== Track 4: Ontologies and Data ===
* Track Chair: Ravi Sharma
This track will cover a range of topics and questions such as:
* What is the difference between an ontology and a mathematical theory?
* Examples of ontologies with large amounts of instance data (e.g., ABox, KG)
** How effective are these ontologies in practical situations?
** Do the ontologies adequately support reasoning processes?
** What trade-offs may have been considered?
* How Ontologies and AI are being used for Science Nobel Prizes?

== Schedule ==
* [[ConferenceCall_2025_01_15|2025_01_15]] Overview Session '''[[GaryBergCross|Gary Berg-Cross]]'''
* [[ConferenceCall_2025_01_22|2025_01_22]] Keynote Address '''[[NicolaGuarino|Nicola Guarino]]'''
* [[ConferenceCall_2025_01_29|2025_01_29]] Track 1 '''[[GiancarloGuizzardi|Giancarlo Guizzardi]]'''
* [[ConferenceCall_2025_02_05|2025_02_05]] Track 1 '''[[MichaelGruninger|Michael Gruninger]]'''
* [[ConferenceCall_2025_02_12|2025_02_12]] Track 1 '''[[BarrySmith|Barry Smith]]'''
* [[ConferenceCall_2025_02_19|2025_02_19]] Track 1 '''[[KenBaclawski|Ken Baclawski]]'''
* [[ConferenceCall_2025_02_26|2025_02_26]] Track 2 '''[[JohnSowa|John Sowa]]'''
* [[ConferenceCall_2025_03_05|2025_03_05]] Track 2 '''Øystein Linnebo'''
* [[ConferenceCall_2025_03_12|2025_03_12]] Track 2 '''Kit Fine'''
* [[ConferenceCall_2025_03_19|2025_03_19]] Synthesis I '''[[KenBaclawski|Ken Baclawski]]'''
* [[ConferenceCall_2025_03_26|2025_03_26]] Track 3 '''[[AndreaWesterinen|Andrea Westerinen]]'''
* [[ConferenceCall_2025_04_02|2025_04_02]] Track 3 '''[[AlicanTüzün]]'''
* [[ConferenceCall_2025_04_09|2025_04_09]] Track 3 TBA
* [[ConferenceCall_2025_04_16|2025_04_16]] Track 4 '''[[RaviSharma|Ravi Sharma]]'''
* [[ConferenceCall_2025_04_23|2025_04_23]] Track 4 '''[[Elisa Kendall - Thematix and OMG ]]'''
* [[ConferenceCall_2025_04_30|2025_04_30]] Track 4 TBA '''No Meeting '''
* [[ConferenceCall_2025_05_07|2025_05_07]] Track 4 '''[[Terry Bollinger - Bottom-Up Time Construction as a Unifying Physics Theme ]]'''
* [[ConferenceCall_2025_05_14|2025_05_14]] Track 4 TBA
* [[ConferenceCall_2025_05_21|2025_05_21]] Synthesis II
* [[ConferenceCall_2025_05_26|2025_05_26]] Communiqué

== Resources ==
* [[OntologySummit2025/ConferenceCallInformation|Conference Call Information]]
* [https://www.youtube.com/channel/UCKK2e8NZ9lyLDBpXY18O_Tg Ontology Summit YouTube Channel]

[[Category:OntologySummit]]
[[Category:OntologySummit2025]]

ConferenceCall 2023 11 08

2023-11-13T16:14:02Z

AndreaWesterinen: /* Agenda */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Broader thoughts]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::8 Nov 2023 17:00 GMT]]
|-
| 9:00am PST/12:00pm EST
|-
| 5:00pm GMT/6:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', strategist and blogger at [http://ailev.lievjournal.ru Laboratory Log]
** ''Hybrid Reasoning, the Scope of Knowledge, and What Is Beyond Ontologies?''
** This talk discusses styles of definitions for knowledge graphs (KG) combined with large language models (LLMs). The KG architectures and systems are reviewed, taken from Ontolog Forum's 2020 Communique. A framework is proposed for a cognitive architecture using both LLMa and KGs for the evolution of knowledge during 4E (embodied, extended, embedded, enacted) cognition. In this framework, ontologies are understood as answers to the question "What is in the world?" and can be found in representations that vary across a spectrum of formality/rigor. An example is given of the use of ontology engineering training in management, where upper-level ontologies are given to students in the form of informal course texts (with the goal of obtaining a fine-tuned LLM within the "neural networks" of students' brains) coupled with lower-level ontologies that are more formal (such as data schemas for databases and knowledge graphs).
** Anatoly Levenchuk has worked as a strategy consultant for more than 30 years. He helps with vision and strategy definition to many government agencies and large companies. Now he is science head of Aisystant that serves as a school in engineering and management. His first machine learning project was in 1977, first ontology engineering project was in 1980. He is author of several textbooks on systems thinking, methodology, systems engineering, systems management, natural and artificial intelligence, education as "person engineering". His blog "Laboratory Log" http://ailev.lievjournal.ru in Russian has more than 3,000 subscribers.
** [https://bit.ly/3sljmXt Slides]
* '''[[ArunMajumdar|Arun Majumdar]]''' and '''[[JohnSowa|John Sowa]]''', [https://permion.ai/ Permion AI]
** ''Trustworthy Computation: Diagrammatic Reasoning With and About LLMs''
** Large Language Models (LLMs) were designed for machine translation (MT). Although LLM methods cannot do any reasoning by themselves, they can often find and apply reasoning patterns that they find in the vast resources of the WWW. For common problems, they frequently find a correct solution. For more complex problems, they may construct a solution that is partially correct for some applications, but disastrously wrong or even hallucinogenic for others. Systems developed by Permion use LLMs for what they do best. But they combine them with precise and trusted methods of diagrammatic reasoning based on conceptual graphs (CGs). They take advantage of the full range of technology developed by 60+ years of AI, computer science, and computational linguistics. For any application, Permion methods derive an ontology tailored to the policies, rules, and specifications of the project or business. All programs and results they produce are guaranteed to be consistent with that ontology.
** [https://bit.ly/464bRlF John's Slides]
** [https://bit.ly/475rpH1 Arun's Slides]
* [https://bit.ly/3QMfqZj Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 8 November 2023'''
* Start Time: 9:00am PST / 12:00pm EST / 6:00pm CET / 5:00pm GMT / 1700 UTC
** Note that Daylight Saving Time has ended in Europe, Canada and the US.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=8&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[JohnSowa|John Sowa]]
* [[ArunMajumdar|Arun Majumdar]]
* [[AnatolyLevenchuk|Anatoly Levenchuk]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* [[AmitSheth|Amit Sheth]]
* [[DavidEddy|David Eddy]]
* [[DaleFitch|Dale Fitch]]
* Nancy Wiegand
* [[PeteRivett|Pete Rivett]]
* [[VictorAgroskin|Victor Agroskin]]
* MZeng
* Tony Cohn (Leeds)

Complete list of participants was not captured

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: When I think about models, it is not necessarily graph or even video but something like a kind of vision mind-based understanding and yes I express it often in language or express math using the language also.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Yes, any form. But if it (model) has patterns then you can consider it as text -- semiotics tell us "all is texts of patterns as letters". Pattern languages are about behavior as a text (chain of patterns). Mathematics is about patterns too.

* [[MichaelDeBellis|Michael DeBellis]]: I don't understand what the speaker means by quantum/digital memory.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: It is about exact copy of information, information is about difference. 1 bit is a result of measurement. To evolution can proceed, you need genes as exactly copied information about results of previous evolution steps. If you have not digital/quantum/discrete form for exact copying, you cannot accumulate knowledge, error in analog form will be prohibitive for evolution.

* [[RaviSharma|Ravi Sharma]]: Second bullet can you expand how cognition and quantum memory are understood?
** [[MikeBennett|Mike Bennett]]: That is the 2nd bullet of Slide 3 (for later reference in the Q&A)
** [[AndreaWesterinen|Andrea Westerinen]]: This is discussed at the end of the talk, in the Q&A

* [[RaviSharma|Ravi Sharma]]: Generative and interactive models, how will these be integrated at different levels?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Generative and discriminative (not interactive) models. You can ask about any models, but get different types of answers: possible worlds descriptions from generative models and classification labels from discriminative models. With many nuances, sure )))
** [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk I think your distinction between "interactive" models is very important!!! See for example the interactionist paradigm of computing by Peter Wegner and Dinah Goldin.

* [[RaviSharma|Ravi Sharma]]: Anatoly would you call L4 Contemplative?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: No. All models about activity (enactive cognition, activity is everywhere).

* [[RaviSharma|Ravi Sharma]]: Generation Differentiation is described, how do you integrate to get all facts together that means knowledge?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: I simply deny all guesses that give problems in inference. All survived guesses are integrated (i.e., not give errors in inference).

* [[RaviSharma|Ravi Sharma]]: Last Q for Anatoly, how do you introduce value systems in these cognitive architectures other than through social media etc.

* [[DouglasMiles|Douglas Miles]]: The most interesting and confusing thing about LLMs is we have no idea how to teach them any new skills... other than: We fire a hose of data and text at them and just pray

* [[AmitSheth|Amit Sheth]]: About passing Turing Test: Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as You May Think -- Introducing AI Detectability Index, https://lnkd.in/gr6cizEZ
** [[ArunMajumdar|Arun Majumdar]]: The distinction between "human" and machine is being attacked on stylometric principles but this is very hard as computer GPT outputs are mixed with human inputs ...

* [[RaviSharma|Ravi Sharma]]: What efforts are likely to succeed in making ChatGPT more accurate?
** [[ArunMajumdar|Arun Majumdar]]: @Ravi Sharma so our approach is to use Conceptual Graphs as our formal knowledge graph approach to creating a "surrogate" model to drive the LLM/GPTs.

* [[ArunMajumdar|Arun Majumdar]]: We (John and I) are not focused on the detection problem because that is something that is not related to our primary focus in knowledge graphs (Conceptual Graphs) as a formalism for symbolic AI with Generative AI.
** [[ArunMajumdar|Arun Majumdar]]: I can share screenshots that compares what we do to what others do. Basically, zero-hallucinations. We can look at a simple hallucination problem later after John's talk if needed. Ken has my screenshots.

* [[RaviSharma|Ravi Sharma]]: Is math not derived from metaphysics?
** [[ArunMajumdar|Arun Majumdar]]: Math is one of the great abstractions of human kind that is a formal but extremely open field for creativity. The key is that creativity in math is often understated. Closed World Models (CWMs) which include LLMs are not capable of creating outside something that they have been input (in their ML training, even with human reinforcement feedback learning).
** [[ArunMajumdar|Arun Majumdar]]: Tuning a GPT or LLM is still a closed world model.
** [[ArunMajumdar|Arun Majumdar]]: Semiotic models differ - the semiotics of Saussure are different to Pierce. One is dyadic, and the other is triadic. So the semiotics matter. These distinctions are not involved in GPT/LLM constructs. However, this important distinction may be approachable in the future.
** [[ArunMajumdar|Arun Majumdar]]: The patterns in mathematics are not probabilistic. They are driven by rational principles. So these are different things.
** [[ArunMajumdar|Arun Majumdar]]: You can use probability, like you can use a million monkeys typing with probabilistic bias to get some pattern candidates but there is no deep insight or principle driving these outputs.

* [[RaviSharma|Ravi Sharma]]: LLM and cognitive scientists and cog-memory well explained.

* [[JanetSinger|Janet Singer]]: Question for Anatoly re 4E cognition. Shift to 4E seems to be the key for appreciating and taking advantage of human-machine teaming opportunities post LLM revolution. But do you see 4E as applicable to software AIs separated from humans? Or just to robots (arguably embodied and embedded if in limited sense)? Or is it that we are the 4E extensions of the software AIs?

* [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk Behavior is not truly just a random process. There is "intention" in living beings - such as the need to survive and thrive. Stochastics are a way to study behavior, to mimic some behaviors but the complexity of goal-driven and intentful behavior is actually much more complex than probabilities.

* [[ArunMajumdar|Arun Majumdar]]: Permion uses tensor mathematics but in the logical formalisms so that predications and symbolic logic has a mapping to/from from the tensorial structures. There are several emerging developments including, for example, the exploration of alternatives algebras such as Clifford or Geometric algebras beyond the conventional real valued Gibbs algebras used nowadays.

* [[DavidEddy|David Eddy]]: 1.5M LoC is NOT a big system

* [[DouglasMiles|Douglas Miles]]: The earth may not be round enough to have a completely circular cord
** [[DouglasMiles|Douglas Miles]]: It takes a person that had experience beyond words to answer that
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a thought experiment discussed by John Sowa - about a circular cord around the earth's equator, and adding 1 yard to it. It raises the cord to approx 6" above the ground.

* [[MikeBennett|Mike Bennett]]: For those of us not from the US, what is the capital of Alaska? It feels like this is a vital plot point.
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a ChatGPT response in John's/Arun's presentation - that Juneau is the capital of Alaska, but there have been attempts to move the capital to another city. And, the underlying issue is whether this information was in the provided training data/corpus (e.g., could you track provenance?)

* [[RaviSharma|Ravi Sharma]]: Arun Say more about scaffolding. Is it only E and R and detailed parsed later, but using what rules?
** [[ArunMajumdar|Arun Majumdar]]: Scaffolding is based on purely mathematical methods based on the Zipf distribution laws of terminologies and the H-Point of the Zipf distribution and Eigen computation methods to identify the modules of graph structures from the text. Scaffolding provides a language
agnostic foundation.

* [[DaleFitch|Dale Fitch]]: Is your "logic" an ontology?
** [[ArunMajumdar|Arun Majumdar]]: Yes we induce and also human update the "ontology" but the logic is using the ontology

* [[DavidEddy|David Eddy]]: Thank you Ken, John, Arun & Anatoly.

== Resources ==
* [https://bit.ly/3QMfqZj Video Recording]
* [https://youtu.be/G4S-Zrc5qUk YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 25

2023-11-13T15:42:13Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 2]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::25 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Evren Sirin''', Stardog CTO and lead for their new [https://www.stardog.com/categories/voicebox/ Voicebox] offering
** '''Title:''' Stardog Voicebox: LLM-Powered Question Answering with Knowledge Graphs
** '''Abstract:''' Large Language Models (LLMs) and Generative AI technologies have caused a shift in all areas of information technology but especially for question answering use cases. Leveraging LLMs for question answering can help fully democratize enterprise analytics and data access. However, using LLMs with enterprise data bring significant challenges around security, privacy, accuracy, and explainabilty. In this talk we will present Stardog [https://www.stardog.com/categories/voicebox/ Voicebox] which leverages an open-source foundational LLM to build, manage, and query knowledge graphs using ordinary language. The answers to user questions directly come from the knowledge graph providing complete traceability and access control. Stardog Voicebox combines statistical reasoning in the form of LLMs with logical reasoning in knowledge graphs providing a powerful hybrid reasoning system with a natural language interface.
** [https://bit.ly/3MeGrSy Slides]
* '''Yuan He''', Key contributor to [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a package for ontology engineering with deep learning
** '''Title:''' DeepOnto: A Python Package for Ontology Engineering with Deep Learning and Language Models
** '''Abstract:''' Integrating deep learning techniques, particularly language models (LMs), with knowledge representations like ontologies has raised widespread attention, urging the need for a platform that supports both paradigms. However, deep learning frameworks like PyTorch and Tensorflow are predominantly developed for Python programming, while widely-used ontology APIs, such as the OWL API and Jena, are primarily Java-based. To facilitate seamless integration of these frameworks and APIs, we present [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a Python package designed for ontology engineering with deep learning. The package encompasses a core ontology processing module founded on the widely-recognized and reliable OWL API, encapsulating its fundamental features in a more “Pythonic” manner and extending its capabilities to incorporate other essential components including reasoning, verbalization, normalization, projection, taxonomy, and more. Building on this module, DeepOnto offers a suite of tools, resources, and algorithms that support various ontology engineering tasks, such as ontology alignment and completion, by harnessing deep learning methods, primarily pre-trained LMs.
** [https://bit.ly/46VaEOu Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 25 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=25&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* Evren Sirin
* Yuan He
* Jiaoyan Chen
* Hang Dong
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* Riley Moher
* [[ToddSchneider|Todd Schneider]]
* [[PeteRivett|Pete Rivett]]
* [[DouglasMiles|Douglas Miles]]
* [[JohnSowa|John Sowa]]
* [[BartGajderowicz|Bart Gajderowicz]]
* E S
* [[GaryBergCross|Gary Berg-Cross]]
* silke
* James Logan
* [[AlexShkotin|Alex Shoktin]]
* Anders Tell
* [[AnatolyLevenchuk|Anatoly Levenchuk]]
* Nancy Wiegand
* Stefan De Giorgis
* [[JohnSowa|John Sowa]]

Complete list of participants was not captured

== Discussion ==
* (Question before chat recording) How is reasoning supported in Stardog?
** [[AndreaWesterinen|Andrea Westerinen]]: Stardog includes a reasoner. (One of the precursors to Stardog was the Pellet reasoner, which could be used in Protege.)

* [[RaviSharma|Ravi Sharma]]: How does stardog distinguish the contention between datastores APIs Vs App API's differences? Namely distinguishing any discrepencies?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure that I understand your question. The datastore APIs are REST based.

* [[AndreaWesterinen|Andrea Westerinen]]: Making SPARQL more user friendly (NLP -> SPARQL) is valuable.
** [[MichaelDeBellis|Michael DeBellis]]: Andrea, can you say more about NLP -> SPARQL? Is this a new spec, a book,??? Have a link?
** [[MichaelDeBellis|Michael DeBellis]]: Sorry, I misread your comment. I thought you said "is available" not "is valuable" and thought it was some new paper or spec. Anyway, yes I agree this is critical to making Semantic Web get traction in the real world and shouldn't be too difficult. Taking natural language and generating SPARQL from it.
** [[AndreaWesterinen|Andrea Westerinen]]: This is what Evren will talk about ... Voicebox.

* [[AndreaWesterinen|Andrea Westerinen]]: Knowledge comes from KGs only, but NLP is for user-assistance.

* [[AndreaWesterinen|Andrea Westerinen]]: Evren is discussing one example of Stardog knowledge kits - related to beers, ingredients and customers.
** [[AndreaWesterinen|Andrea Westerinen]]: Plus there are inference rules that can also be used in a query.
** Riley Moher: So are we generating a new relation whose sort constraints are determined based on semantic similarity of existing relations?
** [[AndreaWesterinen|Andrea Westerinen]]: You can do a PATH analysis, but this is a discussion of an inference rule that I believe is pre-defined.

* [[ToddSchneider|Todd Schneider]]: When asking for a customer's supplier, what are they supplying?
** [[AndreaWesterinen|Andrea Westerinen]]: Any ingredient used in a product purchased by a customer
** [[AndreaWesterinen|Andrea Westerinen]]: Kind of weird, but I think that the point is using inference to help with NLP translation

* [[MichaelDeBellis|Michael DeBellis]]: I would like to get more detail on how you integrated vector DB with triplestore

* [[PeteRivett|Pete Rivett]]: What's the language used to express the rules?
** [[ToddSchneider|Todd Schneider]]: The ‘rules’ look like SWRL rules (i.e. Horn clauses).
** [[AndreaWesterinen|Andrea Westerinen]]: That is what Stardog uses. Yes.

* [[DouglasMiles|Douglas Miles]]: What opensrc LLMs did you find adequately trained?
** [[AndreaWesterinen|Andrea Westerinen]]: Evren is reporting use of MPT-30B trained by MosaicML/Databricks
** [[MichaelDeBellis|Michael DeBellis]]: For open source LLMs the best place to look IMO is: https://huggingface.co/
** [[DouglasMiles|Douglas Miles]]: So you tried several openSrc Models and only one was barely adequate?!
** Evren Sirin: This was based on an analysis of smaller models (7B). And, this is an ongoing process.
** [[DouglasMiles|Douglas Miles]]: I am doing some NL<->CLIF and i so far only found ChatGPT-4 adequate (CahtGPT-3.5 was absolutely a waste of time).. This is good news if MPT-30B is worth a try for this.
** Riley Moher: Very interested in NL <-> CLIF , what is the nature of the work?
** Evren Sirin: LLM stage is changing rapidly. There are some newcomers like Mistral that is promising. We’ve gotten comparable results with Llama 2 as well.
** [[DouglasMiles|Douglas Miles]]: I wont be surpised if Llama X or whichever will be as good as ChatGPT-4 w/in a short time from now
** [[DouglasMiles|Douglas Miles]]: Just a small amount of seeing how good it is showing someone they probably want to use CLIF over the overly popular ARM for KR
** [[DouglasMiles|Douglas Miles]]: I had lots of good experience with CLIF.. It is at least expressive enough for English... whereas ARM seems not to be
** [[JohnSowa|John Sowa]]: CLIF is much easier to map to and from English than OWL.
** Riley Moher: More expressive for sure
** [[JohnSowa|John Sowa]]: CLIF includes full FOL plus a version of HOL.

* [[ToddSchneider|Todd Schneider]]: The results of computing an inference may or may not be materialized.
** Evren Sirin: Yes, in theory inferences can be materialized but in Stardog we only support query-time inferencing by rewriting the user query and executing with a Datalog engine.

* [[PeteRivett|Pete Rivett]]: Have you considered using an enterprise-specific KG to train a specific LLM?
** Evren Sirin: Yes, this is definitely on our roadmap. We are starting with a general-purpose LLM that can be used with any KG but more domain-specific fine-tuning will certainly improve results.

* [[RaviSharma|Ravi Sharma]]: Example of Vector Embeddings?
** Yuan He: SentenceBERT / FAISS, for example.
** Evren Sirin: We use MiniLM which is a small language model (or more correctly a sentence transformer) for creating vectors from text.

* [[BartGajderowicz|Bart Gajderowicz]]: Can you expand on the traceability aspect?
** Evren Sirin: I tried to showcase this at the end a little. We can see the query used, the data sources and data elements that contributed to the answer, etc. Traceability and explainability is still very low-level (requires RDF and SPARQL knowledge) but that’s something we plan to tackle later.

* [[AndreaWesterinen|Andrea Westerinen]]: What happens if the concept is not known to the KG?
** Riley Moher: What if the natural language query cannot be expressed with SPARQL?
** Evren Sirin: At this stage we simple say question cannot be answered. Our goal is to use the conversational aspects to clarify the question and/or clarify to the user that graph does not contain relevant information.

* E S: @Evren Sirin Supply Chain usage demonstration is very useful and I think applicable in business. What are the required specs for hardware for Stardog? Would you please share the estimated costs for monthly fixed costs to maintain the whole system, ie regardless of customers usage? Thank you.
** Evren Sirin: There are lots of different considerations that would go into hardware specs. We typically suggest people to start with out hosted option that has a Free tier:https://www.stardog.com/stardog-cloud/ For on-prem deployment, there is capacity planning discussion here https://docs.stardog.com/operating-stardog/server-administration/capacity-planning

* Jiaoyan Chen: We are happy to answer questions on DeepOnto in the ChatBox.

* [[RaviSharma|Ravi Sharma]]: Yuan Great start, on ontology engineering

* [[AndreaWesterinen|Andrea Westerinen]]: Where was the FoodProduct rule defined?

* [[ToddSchneider|Todd Schneider]]: Explainability is a huge benefit over LLMs.

* [[RaviSharma|Ravi Sharma]]: I see verification and logical merging, embedding etc.

* [[ToddSchneider|Todd Schneider]]: Yuan, please explain ‘subsumption restructuring’.1
* Jiaoyan Chen: I do not fully capture in which slide this phrase happens. If it is in ontology-to-graph, it refers to our work that extracting a class hierarchy from the original ontology. There is special case: A \equiv B \conjunciton C, we will have A \subclassof B and A \subclassof C; this is to avoid placing A under owl:Thing, if we just consider the declared subsumptions of named classes for building the hierarchy. This is also the strategy of Protege for pressing the class hierarchy.

* [[RaviSharma|Ravi Sharma]]: What happens if semantics from same entities have differences?
* Hang Dong: One idea is to do automated concept discovery and insertion (from texts for example). We are still exploring to implement these in DeepOnto. One recent work https://arxiv.org/abs/2306.14704 and https://arxiv.org/abs/2302.07189

* [[GaryBergCross|Gary Berg-Cross]]: @Yuan Can you say a bit more about how alignment between ontologies is done?
* Jiaoyan Chen : Briefly, it fine-tunes a BERT-based binary classifier with synonyms from the ontologies to be aligned, uses the classifier to predict candidate equivalent class pairs with class labels, combines the prediction scores with lexical matching scores, and finally uses logical reasoning for consistency checking and repair (using a repairing algorithm our group developed before).
** [[RaviSharma|Ravi Sharma]]: In that case how is alignment taken care of?

* [[MichaelDeBellis|Michael DeBellis]]: Both of these presentations were excellent! Very useful info.

* [[RaviSharma|Ravi Sharma]]: Can you tell more about OAEI?
** Jiaoyan Chen: https://www.cs.ox.ac.uk/isg/projects/ConCur/oaei/2023/index.html
** Hang Dong: It is an onto matching benchmarking activity running for many years.
** Jiaoyan Chen: We placed a new Bio-ML track in OAEI which has been made for over a decade.
** Jiaoyan Chen: Our new Bio-ML track was place in 2022, and is continuing in 2023. This track is especially developed for ML-based OM systems.

* [[JohnSowa|John Sowa]]: If you use LLMs to do translation from English to CLIF, the mapping is simpler and more successful.

* silke: Can you please give a short explanation of logic repair and how it is implemented? Thanks!
** Jiaoyan Chen: Yes. We get mappings and their scores. Briefly, the repair algorithm merges the mappings and the ontologies to infer whether they are consistent. If not, it tries to remove some mappings with lowest scores, and see whether the remaining mappings + the ontologies are consistent. If yes, it stops. This procedure is iterative. The reasoning is approximated using Propositional logics.
** Jiaoyan Chen: More details are here: https://ceur-ws.org/Vol-1014/paper_63.pdf
** silke: Thank you so much, very helpful!

* [[JohnSowa|John Sowa]]: For any mapping from NL to any other notation, you need an "echo".
** [[JohnSowa|John Sowa]]: Whenever you type anything in English, the system should produce an echo in English to show exactly how your input was interpreted.
** [[DouglasMiles|Douglas Miles]]: I have asked it to translate the CLIF back to English and then told it to tell me if the original English matches it response and to modify the English->CLIF .. tis 2nd round produces much better results
** [[JohnSowa|John Sowa]]: If the echo is not what you wanted, you can revise your question.

* James LOGAN: Can DeepOnto create axioms that seem to always hold true in some domain from a text corpus?
** Jiaoyan Chen: Not yet. We now are trying to extract new concepts from text and insert them into the ontology (there are some ongoing works: https://arxiv.org/abs/2306.14704. We haven’t consider axioms, but only concepts. It’s a good idea for the future extension.

* [[ToddSchneider|Todd Schneider]]: Please explain ‘subsumption restructuring’.
** Yuan He: We introduced subsumption axioms between parents and children concepts of a concept target for removal.
** James LOGAN: It seems this would require jumping to conclusions or having a way to close the world

* [[MikeBennett|Mike Bennett]]: Yuan is the ontology alignment etc. dependent on any specific TLO? Or can different TLOs be used as the basis for this?
** Jiaoyan Chen: I don’t know what’s TLO, but I think not …
** [[MikeBennett|Mike Bennett]]: Top Level Ontology
** Yuan He: Just depend on the input ontologies is sufficient.
** Jiaoyan Chen: No, it does not
** [[ToddSchneider|Todd Schneider]]: ‘Top Level Ontology’ equivalent to ‘Foundational Ontology’

* E S: Is there any accuracy problem in building KG with other languages than English?
** Jiaoyan Chen: DeepOnto currently is tested only for English ontologies
** [[AndreaWesterinen|Andrea Westerinen]]: You need the appropriate training for LLMs. So, you have translation and translation errors.

== Resources ==
* [https://bit.ly/3MeGrSy Evrin Sirin Slides]
* [https://bit.ly/46VaEOu Yuan He Slides]
* [https://bit.ly/3Q3o0Be Video Recording]
* [https://youtu.be/OYJWARcH5B0 YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 01

2023-11-12T23:30:58Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Demos of information extraction via hybrid systems]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::1 Nov
2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* [[AndreaWesterinen|Andrea Westerinen]], Creator of [https://github.com/ontoinsights/deep_narrative_analysis DNA, Deep Narrative Analysis]
** '''Title:''' Populating Knowledge Graphs: The Confluence of Ontology and Large Language Models
** '''Abstract:''' Ontology-based Knowledge Graphs (KGs) stand at the forefront of semantic data representation, providing structured views of the data in complex domains. Traditionally, populating these KGs from unstructured text involved convoluted natural language analyses and custom code, but the environment has changed with the use of Large Language Models (LLMs). This talk explores one use case - the population of a KG from news articles. The evolution of the DNA application from employing spaCy APIs to OpenAI is described, and the current (open-source) implementation discussed. Implementation issues such as sourcing the data, LLM prompts, mapping the LLM responses onto the ontology, and populating the knowledge graph are overviewed.
** [https://bit.ly/49tTuJY Slides]
* [[PrasadYalamanchi|Prasad Yalamanchi]], [https://leadsemantics.com/ Lead Semantics] CTO
** '''Title:''' Harvest Knowledge From Language - Harness the power of Large Language Models and Semantic Technology
** '''Abstract:''' Language (both text and voice) holds much of the accessible knowledge to humans. It is also the best store of the collective human knowledge. Historically, accessing this knowledge, was manual and up until recent times has progressed to varying degrees of semi-automated methods! But, with the advent of Language Models and particularly Large Language Models in the last couple of years, a fully automated access to knowledge carried in language is now becoming a reality! TextDistil, the software product from Lead Semantics applies LLMs and Ontologies to extract computable knowledge in the form of RDF triples from Text.
** [https://bit.ly/3tY3niI Slides]
* [https://bit.ly/45Z1fUY Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 1 November 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 5:00pm CET / 4:00pm GMT / 1600 UTC
** Note that Daylight Saving Time has ended in Europe but not in the US or Canada.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=1&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[PrasadYalamanchi|Prasad Yalamanchi]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[JohnSowa|John Sowa]]
* [[BartGajderowicz|Bart Gajderowicz]]
* [[GaryBergCross|Gary Berg-Cross]]
* Zefi Kavvadia
* Dan (Telicent)
* Helena (Telicent)
* [[MarkUnderwood|Mark Underwood]]
* Sundos Al Subhi
* Jeff
* James Logan
* Gian Piero Zarri
* Mariusz Bronowicki
* [[RamSriram|Ram Sriram]]
* Jim Rhyne
* [[MarkRessler|Mark Ressler]]
* Robin McEntire
* Ibrahim Gh
* Matt Turner
* Asiyah Yu Lin
* [[MarkFox|Mark Fox]]
* [[AlexShkotin|Alex Shkotin]]
* [[VictorAgroskin|Victor Agroskin]]

Complete list of participants was not captured

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: What is the normal accuracy and does the accuracy of triples vary by language?
** [[RaviSharma|Ravi Sharma]]: Namely would it just depend on the language only or on domain concept would affect the KG?

* [[MichaelDeBellis|Michael DeBellis]]: Is the ontology COMPLETELY created from the Corpus, or do you start from a foundation ontology and extend it based on the Corpus?

* [[RaviSharma|Ravi Sharma]]: You implied preprocessing and semantic understanding by humans before the KG is generated? how much effort is it?

* [[RaviSharma|Ravi Sharma]]: Is there a possibility to reduce the duration by compromising the accuracy somewhat?

* [[SusanneVejdemo|Susanne Vejdemo]]: I understand we have a KG constructed from the unstructured docs. And then there’s translation of your query to triples? I am a bit uncertain where the LLM comes into this?
** [[JanetSinger|Janet Singer]]: My question as well — how exactly does the LLM come in?
** [[AndreaWesterinen|Andrea Westerinen]]: I will address this in my presentation, but can’t answer for Prasad.
** [[PrasadYalamanchi|Prasad Yalamanchi]]: LLM is coming in multiple places in the TextDistil pipeline. Once at the final summary string of the result items. It comes in building the KG, as well
** [[BartGajderowicz|Bart Gajderowicz]]: Prasad, what does the LLM do in building the KG?

* [[RaviSharma|Ravi Sharma]]: Can it show the visuals during progress such as the KG?

* [[DouglasMiles|Douglas Miles]]: Chatgpt-3.5 in some instances can work well enough to be used ovber ChatGPT-4 ?
** [[AndreaWesterinen|Andrea Westerinen]]: It MAY, but I have found profound differences. Linguistic analysis is much better in 4
** [[DouglasMiles|Douglas Miles]]: ChatGPT-3.5 can be so much faster with its return results.. I've considered running both to see when 3.5 was sufficient.. admnittely mostly it isn't.. but "convert this to owl" often is "acceptable"
** [[AndreaWesterinen|Andrea Westerinen]]: @Douglas Miles Not sure that I agree about acceptability.
** [[DouglasMiles|Douglas Miles]]: ok true.. I mean 3.5 cant even begin to convert to CLIF of CycL .. whereas it at least tried with RDF/OWL

* [[RaviSharma|Ravi Sharma]]: If there are no human interventions, how much is the KG affected?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi, that's my question as well. IMO it is usually better to have a person in the loop because creating a well designed ontology completely from a Corpus seems like the resulting ontology may not be well designed. That's why I asked the question about starting from a basic ontology and then extending that ontology, rather than creating the entire ontology from scratch

* [[DouglasMiles|Douglas Miles]]: Prasad presentation was very awesome!

* [[RaviSharma|Ravi Sharma]]: Prasad - If you do two such exercises, is the result the same/repeatable?
** [[PrasadYalamanchi|Prasad Yalamanchi]]: There is a language interpreatation to map the query string to the Ontology.
** [[PrasadYalamanchi|Prasad Yalamanchi]]: SO, if two queries (exercises as you mentioned) result in the same interpretation, then the final answers will be the same

* [[DouglasMiles|Douglas Miles]]: Something that impresses me and is unique about Andrea's work (even year or two ago.. ) ... She actually supports full modality representations in RDF-ish languages.. Stuff that normally I would only dare to use CLIF to represent!

* [[RaviSharma|Ravi Sharma]]: Are there similarities to the rhetoric possibilities of metaphor, context, explanation, etc to improve your results?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure what is being asked. I am exposing the use of rhetorical devices to help readers understand how the text might be affecting their interpretations of it.

* [[RaviSharma|Ravi Sharma]]: Andrea does ML or AI enter this exercise? And results you showed, if so where?
* [[RaviSharma|Ravi Sharma]]: I mean what ML and learning sets were used in OpenAI
** [[AndreaWesterinen|Andrea Westerinen]]: OpenAI's complete technology stack is not disclosed but their website says "We build our generative models using a technology called deep learning, which leverages large amounts of data to train an AI system to perform a task."

* [[DouglasMiles|Douglas Miles]]: To help answer how LLMs can be useful in translation: https://chat.openai.com/share/039d72c3-8432-48d1-98b8-63e15614bbef

* [[MarkUnderwood|Mark Underwood]]: Excellent presentations and important work for the ontology community

* Sundos Al Subhi: Thank you all!! Great information.

* [[JanetSinger|Janet Singer]]: Excellent presentations — Looking forward to seeing these ideas integrated in the future session(s)

* [[DouglasMiles|Douglas Miles]]: (there is no question that the KR Andrea is doing is rock solid!) Here is my question though: Are any of the RDF reasoners good enough to do the reasoning/query that Andrea expects?
** [[AndreaWesterinen|Andrea Westerinen]]: @Douglas Miles Yes, I use Stardog. Also allows use of Voicebox which encode NL queries in SPARQL!

* [[DouglasMiles|Douglas Miles]]: Thank you that was great!

* Dan (Telicent): Thank you for your presentations 🙂

* Zefi Kavvadia : thank you!

* Mariusz (Telicent) : Thank you.

== Resources ==
* [https://bit.ly/45Z1fUY Video Recording]
* [https://youtu.be/h9Rl6nZ50Ls YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_01]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_01]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 01

2023-11-12T23:26:42Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Demos of information extraction via hybrid systems]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::1 Nov
2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* [[AndreaWesterinen|Andrea Westerinen]], Creator of [https://github.com/ontoinsights/deep_narrative_analysis DNA, Deep Narrative Analysis]
** '''Title:''' Populating Knowledge Graphs: The Confluence of Ontology and Large Language Models
** '''Abstract:''' Ontology-based Knowledge Graphs (KGs) stand at the forefront of semantic data representation, providing structured views of the data in complex domains. Traditionally, populating these KGs from unstructured text involved convoluted natural language analyses and custom code, but the environment has changed with the use of Large Language Models (LLMs). This talk explores one use case - the population of a KG from news articles. The evolution of the DNA application from employing spaCy APIs to OpenAI is described, and the current (open-source) implementation discussed. Implementation issues such as sourcing the data, LLM prompts, mapping the LLM responses onto the ontology, and populating the knowledge graph are overviewed.
** [https://bit.ly/49tTuJY Slides]
* [[PrasadYalamanchi|Prasad Yalamanchi]], [https://leadsemantics.com/ Lead Semantics] CTO
** '''Title:''' Harvest Knowledge From Language - Harness the power of Large Language Models and Semantic Technology
** '''Abstract:''' Language (both text and voice) holds much of the accessible knowledge to humans. It is also the best store of the collective human knowledge. Historically, accessing this knowledge, was manual and up until recent times has progressed to varying degrees of semi-automated methods! But, with the advent of Language Models and particularly Large Language Models in the last couple of years, a fully automated access to knowledge carried in language is now becoming a reality! TextDistil, the software product from Lead Semantics applies LLMs and Ontologies to extract computable knowledge in the form of RDF triples from Text.
** [https://bit.ly/3tY3niI Slides]
* [https://bit.ly/45Z1fUY Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 1 November 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 5:00pm CET / 4:00pm GMT / 1600 UTC
** Note that Daylight Saving Time has ended in Europe but not in the US or Canada.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=1&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[PrasadYalamanchi|Prasad Yalamanchi]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[JohnSowa|John Sowa]]
* [[BartGajderowicz|Bart Gajderowicz]]
* [[GaryBergCross|Gary Berg-Cross]]
* Zefi Kavvadia
* ariusz (Telicent)
* Dan (Telicent)
* [[MarkUnderwood|Mark Underwood]]
* Sundos Al Subhi
* Jeff
* James Logan
* Gian Piero Zarri
* Mariusz Bronowicki
* [[RamSriram|Ram Sriram]]
* Jim Rhyne

Complete list of participants was not captured

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: What is the normal accuracy and does the accuracy of triples vary by language?
** [[RaviSharma|Ravi Sharma]]: Namely would it just depend on the language only or on domain concept would affect the KG?

* [[MichaelDeBellis|Michael DeBellis]]: Is the ontology COMPLETELY created from the Corpus, or do you start from a foundation ontology and extend it based on the Corpus?

* [[RaviSharma|Ravi Sharma]]: You implied preprocessing and semantic understanding by humans before the KG is generated? how much effort is it?

* [[RaviSharma|Ravi Sharma]]: Is there a possibility to reduce the duration by compromising the accuracy somewhat?

* [[SusanneVejdemo|Susanne Vejdemo]]: I understand we have a KG constructed from the unstructured docs. And then there’s translation of your query to triples? I am a bit uncertain where the LLM comes into this?
** [[JanetSinger|Janet Singer]]: My question as well — how exactly does the LLM come in?
** [[AndreaWesterinen|Andrea Westerinen]]: I will address this in my presentation, but can’t answer for Prasad.
** [[PrasadYalamanchi|Prasad Yalamanchi]]: LLM is coming in multiple places in the TextDistil pipeline. Once at the final summary string of the result items. It comes in building the KG, as well
** [[BartGajderowicz|Bart Gajderowicz]]: Prasad, what does the LLM do in building the KG?

* [[RaviSharma|Ravi Sharma]]: Can it show the visuals during progress such as the KG?

* [[DouglasMiles|Douglas Miles]]: Chatgpt-3.5 in some instances can work well enough to be used ovber ChatGPT-4 ?
** [[AndreaWesterinen|Andrea Westerinen]]: It MAY, but I have found profound differences. Linguistic analysis is much better in 4
** [[DouglasMiles|Douglas Miles]]: ChatGPT-3.5 can be so much faster with its return results.. I've considered running both to see when 3.5 was sufficient.. admnittely mostly it isn't.. but "convert this to owl" often is "acceptable"
** [[AndreaWesterinen|Andrea Westerinen]]: @Douglas Miles Not sure that I agree about acceptability.
** [[DouglasMiles|Douglas Miles]]: ok true.. I mean 3.5 cant even begin to convert to CLIF of CycL .. whereas it at least tried with RDF/OWL

* [[RaviSharma|Ravi Sharma]]: If there are no human interventions, how much is the KG affected?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi, that's my question as well. IMO it is usually better to have a person in the loop because creating a well designed ontology completely from a Corpus seems like the resulting ontology may not be well designed. That's why I asked the question about starting from a basic ontology and then extending that ontology, rather than creating the entire ontology from scratch

* [[DouglasMiles|Douglas Miles]]: Prasad presentation was very awesome!

* [[RaviSharma|Ravi Sharma]]: Prasad - If you do two such exercises, is the result the same/repeatable?
** [[PrasadYalamanchi|Prasad Yalamanchi]]: There is a language interpreatation to map the query string to the Ontology.
** [[PrasadYalamanchi|Prasad Yalamanchi]]: SO, if two queries (exercises as you mentioned) result in the same interpretation, then the final answers will be the same

* [[DouglasMiles|Douglas Miles]]: Something that impresses me and is unique about Andrea's work (even year or two ago.. ) ... She actually supports full modality representations in RDF-ish languages.. Stuff that normally I would only dare to use CLIF to represent!

* [[RaviSharma|Ravi Sharma]]: Are there similarities to the rhetoric possibilities of metaphor, context, explanation, etc to improve your results?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure what is being asked. I am exposing the use of rhetorical devices to help readers understand how the text might be affecting their interpretations of it.

* [[RaviSharma|Ravi Sharma]]: Andrea does ML or AI enter this exercise? And results you showed, if so where?
* [[RaviSharma|Ravi Sharma]]: I mean what ML and learning sets were used in OpenAI
** [[AndreaWesterinen|Andrea Westerinen]]: OpenAI's complete technology stack is not disclosed but their website says "We build our generative models using a technology called deep learning, which leverages large amounts of data to train an AI system to perform a task."

* [[DouglasMiles|Douglas Miles]]: To help answer how LLMs can be useful in translation: https://chat.openai.com/share/039d72c3-8432-48d1-98b8-63e15614bbef

* [[MarkUnderwood|Mark Underwood]]: Excellent presentations and important work for the ontology community

* Sundos Al Subhi: Thank you all!! Great information.

* [[JanetSinger|Janet Singer]]: Excellent presentations — Looking forward to seeing these ideas integrated in the future session(s)

* [[DouglasMiles|Douglas Miles]]: (there is no question that the KR Andrea is doing is rock solid!) Here is my question though: Are any of the RDF reasoners good enough to do the reasoning/query that Andrea expects?
** [[AndreaWesterinen|Andrea Westerinen]]: @Douglas Miles Yes, I use Stardog. Also allows use of Voicebox which encode NL queries in SPARQL!

* [[DouglasMiles|Douglas Miles]]: Thank you that was great!

* Dan (Telicent): Thank you for your presentations 🙂

* Zefi Kavvadia : thank you!

* Mariusz (Telicent) : Thank you.

== Resources ==
* [https://bit.ly/45Z1fUY Video Recording]
* [https://youtu.be/h9Rl6nZ50Ls YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_01]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_01]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 18

2023-11-12T23:24:20Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 1]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::18 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Kurt Cagle''', Author of [https://thecaglereport.com/ The Cagle Report]
** '''Title:''' Complementary Thinking: Language Models, Ontologies and Knowledge Graphs
** '''Abstract:''' With the advent of Retrieval Augmented Generators (RAGs), a more or less standardized workflow has become available for integrating large language models such as ChatGPT with knowledge graphs. This in turn has raised the question about the nature of ontologies associated with LLMs and how knowledge graphs can be structured and queried to make integrated data access possible between the two types of systems. In this talk, Editor and AI Explorer Kurt Cagle of The Cagle Report looks at this process and discusses how they affect both knowledge portals and ontology design.
** [https://bit.ly/3S040lR Slides]
* '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/ LinkedIn])
** '''Title:''' How Ontologies Can Unlock the Potential of Large Language Models for Business
** '''Abstract:''' LLMs have remarkable capabilities; they can craft letters, analyze data, orchestrate workflows, generate code, and much more. Companies such as Google, Apple, Amazon, Meta, and Microsoft are all investing heavily in this technology. Everything indicates that LLMs have enormous disruptive potential. However, there is a problem: they can hallucinate, and for any serious business, that is a deal-breaker. This is where ontologies can come in. In combination with Knowledge Graphs, they can place guardrails around the LLMs, thus allowing organizations to harness the capabilities of LLMs within the framework of a safely controlled ontological structure.
** [https://bit.ly/46A3EH2 Slides]

[https://bit.ly/46DJyvo Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 18 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=18&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
80+ participants

* Kurt Cagle
* Tony Seale
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* Ayya Niyyanika Bhikkhuni
* Anthony Alcaraz
* Alan Morrison
* [[BartGajderowicz|Bart Gajderowicz]]
* Chris Day
* Andrew McCaffrey
* [[GaryBergCross|Gary Berg-Cross]]
* [[JohnSowa|John Sowa]]
* [[MarkUnderwood|Mark Underwood (IS Innovation)]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[SusanneVejdemo|Sus Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* Yishan Liu
* James Logan
* Jason Polis
* Piers Hollott
* [[PennyAnderson|Penny Anderson]]
* Michael Robbins
* [[MichaelDeBellis|Michael DeBellis]]
* Amit Jain
* Cedric Berger
* Andreas Lothe Opdahl
* Harvey King
* Benoit Claise
* Liju Fan
* Larry Swanson
* Justin Lewis

== Discussion ==
* Anthony's OtterPilot: Hi, I'm an AI assistant helping Anthony Alcaraz take notes for this meeting. Follow along the transcript here: https://otter.ai/u/WLIaj2w-OmOEoVVCP5gURhhZHaY?utm_source=va_chat_link_2  You'll also be able to see screenshots of key moments, add highlights, comments, or action items to anything being said, and get an automatic summary after the meeting.
** [[MikeBennett|Mike Bennett]]: Dang! Put on a session on AI and the AIs start showing up!

* [[AndreaWesterinen|Andrea Westerinen]]: I like the quote, “Ontologies are the shapes of information and knowledge”. Also they provide "information for communication".

* [[ToddSchneider|Todd Schneider]]: What is a ‘shape of information’?
** [[AndreaWesterinen|Andrea Westerinen]]: Roles, constraints, relationships for a domain; a local representation

* [[RaviSharma|Ravi Sharma]]: Can LLMs and/or ontologies be used to detect AI artifacts?
** [[AndreaWesterinen|Andrea Westerinen]]: @Ravi Sharma There are some different articles on this, but basically no.
** [[MikeBennett|Mike Bennett]]: Not unless you could write an ontology that defines what it is to be truly human
** Michael Robbins: Or rebuild the web from the bottom up to embed new frameworks for digital identity and content provenance/authenticity (which is what we need to commit to)
** [[AndreaWesterinen|Andrea Westerinen]]: Provenance also aids with detecting bots

* [[MichaelDeBellis|Michael DeBellis]]: Someone asked about agents in relation to Data Mesh (can't find the orig comment) IMO Data Mesh could be implemented in terms of Agents but typically isn't.

* [[RaviSharma|Ravi Sharma]]: Is there one kind or multiple kinds of connectivity in KG as well as in Ontologies?

* [[RaviSharma|Ravi Sharma]]: Container in the sense of domain overlaps especially overlapping vocabs as Venn diagrams?

* Anh: Why is it hard to build LLMs for other languages? Why can't it be replicated easily when translation work (e.g. Facebook, Google Translate, LinkedIn) has already done a somewhat good job?
** [[AndreaWesterinen|Andrea Westerinen]]: It is about the training data.
** [[BartGajderowicz|Bart Gajderowicz]]: [Anh], on longer text, translation looses many of the cultural nuances of a language, and looses context. Also, most training data is in English so most models are trained on English then translated.
** Anh: Thank you, @Andrea Westerinen & @Bart Gajderowicz. Does it meean it'd cost the same to build a new LLMs for a new language?
** [[AndreaWesterinen|Andrea Westerinen]]: I would believe so.
** [[BartGajderowicz|Bart Gajderowicz]]: The volume of English text for training is cheaper, it’s just the web. So I’d imagine finding enough text in your target language would be the biggest cost. Librarians are our friends here 🙂
** [[PennyAnderson|Penny Anderson]]: this context free is difficult when you start thinking of utility of UI
** Anh: Could you please elaborate more on utility of UI? @Penny Anderson
** [[PennyAnderson|Penny Anderson]]: I mean UIs tend to be process driven data-centric is not tightly bound to a particular process that is context free
** [[PennyAnderson|Penny Anderson]]: other than search ?
** Anh: I see. Thanks.
** [[PennyAnderson|Penny Anderson]]: Data providence, Ethical AI ?
** [[AndreaWesterinen|Andrea Westerinen]]: @Penny Anderson Much better declared and reasoned against in KGs.
** [[PennyAnderson|Penny Anderson]]: Zero trust in networks?

* [[RaviSharma|Ravi Sharma]]: Kurt when you talk of box, you are essentially stating in and out of scope items or is there a way of capturing the info outside the box and bringing it in?

* [[AndreaWesterinen|Andrea Westerinen]]: "What is valid for a graph? Shapes and data"

* [[MichaelDeBellis|Michael DeBellis]]: I don't agree that knowledge graphs are a form of data mesh. They are two related but different concepts. Data mesh is essentially an architecture for enterprise data definition and management. Knowledge graphs are a tool that can be used to implement a data mesh. Data mesh IMO brings the philosophy of microservices to data.
** Alan Morrison: Michael, re: microservices to data, should we be thinking of agents as messengers and KGs as the data resource>

* [[RaviSharma|Ravi Sharma]]: What kind of aggregates are these data shapes? Are these only valid for a class of data or can you mix data types in a shape aggregate?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi: in SHACL you define similar things as you do with OWL. E.g., does a property have to have exactly one value, the datatypes of a property, other constraints. The difference is OWL is used for reasoning over large knowledge graphs and uses the Open World Assumption. SHACL is for constraining data so it uses the Closed World Assumption. E.g., you can define ss_number as a property that must have exactly one value in either OWL or SHACL but in OWL you will almost never trigger an error if the restriction isn't satisfied due to OWA. With SHACL you will get error messages due to CWA

* [[ToddSchneider|Todd Schneider]]: LLM reasoning is probabilistic.

* Ayya Niyyanika Bhikkhuni: When weighting a concept is that some sort of credibility score?

* Michael Robbins: LangChains - I’m thinking about what role ConLangs could serve its intermediaries n revolutionizing language modes and NLP. Happy to have a follow-up discussion with anyone who is interested. https://en.wikipedia.org/wiki/Constructed_language

* [[RaviSharma|Ravi Sharma]]: What is the equivalent of Objects in LLM? What are these entities called and can same onto-entity be different in LLM context?
** [[MikeBennett|Mike Bennett]]: Words, I presume?

* [[RaviSharma|Ravi Sharma]]: Kurt thanks for including wonderful valuable background cultural images, these are inspiring. Are you also conveying the there is external (databased) and internal knowledge such as contemplation?

* [[RaviSharma|Ravi Sharma]]: Why are you limiting your examples to RDF why not MOF also?

* [[ToddSchneider|Todd Schneider]]: The important question is not mapping(s). It’s how can well constructed ontologies be used in the ingestion/training of LLMs.

* [[SusanneVejdemo|Sus Vejdemo]]: Question: what are good case studies of KGE and LLM integrations?
** [[AndreaWesterinen|Andrea Westerinen]]: I addressed some of this in the opening session. “Hybrid systems” include both where LLMs help ontologies (actually the Oct 25th session) and where ontologies help LLMs (Oct 4 and Nov 1 sessions).
** [[AndreaWesterinen|Andrea Westerinen]]: Also, Tony is highlighting a GREAT integration.

* [[MichaelDeBellis|Michael DeBellis]]: One question I have is what happens once you load a knowledge graph into an LLM? I know it can be done but once you load say a Turtle file into the LLM then what?
** [[BartGajderowicz|Bart Gajderowicz]]: Generally the LLM builds a small knowledge graph instance inside its memory and you can query it. Ask it to write SPARQL to get some instances, etc. I have not seen it used for large KGs, just small ones.

* Amit Jain: Will the recording be shared with attendees after the summit?
** [[MikeBennett|Mike Bennett]]: Yes, the recording will be uploaded to the session page when it is ready.

* Cedric Berger: Is there any metrics to measure that indeed KG combines with LLM are less hallucinating?
** [[AndreaWesterinen|Andrea Westerinen]]: I will try to provide these in the summary. I have read papers on this as well as blog posts.
** [[BartGajderowicz|Bart Gajderowicz]]: All the work I’ve come across relies on the knowledge graph to provide explicit knowledge. So if you can ground the LLM with a graph, you can verify if the answers the LLM provides are “facts” in the graph

* Kurt Cagle: Once you load in the ontology into a conversation, it will create (to some extent) an LLM conceptual space for that data. Also keep in mind that getting ALL of a knowledge graph via a RAG is usually not feasible.

* [[RaviSharma|Ravi Sharma]]: Tony great diagram for LLM + ontology to improve each other.why is ontology weak in capturing concepts Vs LLM?

* [[MichaelDeBellis|Michael DeBellis]]: One of the most interesting papers I've read is from Lawrence Berkeley Labs on using LLMs to extend an ontology.

* [[RaviSharma|Ravi Sharma]]: How would models such as in physics work with LLM and Ontology loop or cycle that you show, actually ontologies are conceptually richer than KGs alone?
** [[AndreaWesterinen|Andrea Westerinen]]: KGs are the data, ontologies are the concepts … So, it does not seem right to ask about one being richer.
** Michael Robbins: https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its-wolfram-superpowers/

* [[RaviSharma|Ravi Sharma]]: Tony the built-in uncertainty in LLMs gives it extra power to apply to real life probabilistic world?

* [[RaviSharma|Ravi Sharma]]: Tony LLMs and analog and ontology as Quantum? great way. thanks

* [[MikeBennett|Mike Bennett]]: Left brain v right brain is a great way of talking about ontology v LLM. Now you have to create a good corpus collosum.
** [[GaryBergCross|Gary Berg-Cross]]: A better model than left right hemispheres is by layers - old, mid brain (associative) and neo-cortex. They interconnect in many ways and some by the limbic system.
** [[BartGajderowicz|Bart Gajderowicz]]: I like the System 1 / System 2 analogy

* [[AndreaWesterinen|Andrea Westerinen]]: Tony characterized knowledge graphs as discrete and LLMs as continuous.

* Cedric Berger: Aren’t LLMs also kind of discrete as relying on vectors (arrays of numbers) of limited dimensions?
** [[BartGajderowicz|Bart Gajderowicz]]: LLMs are fundamentally probabilistic, not discrete. Some models are trained to provide discrete classifications, but that’s just at the output level. Internally they are probabilistic.
** [[AndreaWesterinen|Andrea Westerinen]]: The discreteness pertains to the encoding. But, does the probabilistic nature of LLMs make it more continuous? I am not sure.
** [[MikeBennett|Mike Bennett]]: The weightings are a number rather than a logical truth value as in KGs
** Kurt Cagle: Even with KGs, you can set up reifications that also set up Bayesians that are again more fuzzy (or at least more stochastic).
** [[MikeBennett|Mike Bennett]]: If you had an ontology with weightings instead of truth values and can train those values, you have a semantic network like a brain/mind.
** Kurt Cagle: It's where I think we're heading. People in the semantic space have known for years that knowledge is fuzzy / fractal, but getting there has always been the rub.
** [[MikeBennett|Mike Bennett]]: I roughed out an idea for this kind of semantic network application back in the 90s.

* Michael Robbins: A great article on vector embeddings: https://kdb.ai/learning-hub/fundamentals/vector-embeddings/ How can we use this for transparency and explainability? Give users confidence intervals (and other potential response options) along with responses?

* [[RaviSharma|Ravi Sharma]]: What are vectors equivalent to in LLM context?

* Harvey King: Are the embeddings stored across 3 layers?
** [[AndreaWesterinen|Andrea Westerinen]]: The embeddings are there, but simplified/reduced.

* Anh: Does KG consider the time stamp of the assertions/objects? Context of my question: could we use it to mark the originality of posts of similar contents to alert plagiarism.
** [[AndreaWesterinen|Andrea Westerinen]]: The KG CAN do this, if it is encoded.

* [[RaviSharma|Ravi Sharma]]: Has anyone done this a roundtrip quality check, learn from LLM and put it in ontologies and the other way around?
** Benoit Claise: In which context/use case?
** [[MichaelDeBellis|Michael DeBellis]]: That paper I posted earlier from Lawrence Berkeley Labs used LLM to extend an ontology but just went in one direction, expanding the ontology not changing the LLM.

* [[RaviSharma|Ravi Sharma]]: Can you do reasoning on same concept in both to differentiate their respective strengths and weakness?

* [[RaviSharma|Ravi Sharma]]: Tony your tree or chain of thought are great ways

* [[RaviSharma|Ravi Sharma]]: Tony and Kurt Can you address feature space Vs training set learning approaches?

* Liju Fan: Why are the relations in the ontology explicit?
** [[AndreaWesterinen|Andrea Westerinen]]: Relations are defined, and they can be inferred, but this is the essence of ontology. Ontologies are “open world” but do need relationships.
** [[MichaelDeBellis|Michael DeBellis]]: Because in the ontology you (usually manually) create the relations. In an LLM the relations are inferred by the ML algorithm and usually can't be manually changed
** Liju Fan: It seems there is a need to be able to rename LLM inferred relations for them to be human-understandable and practically useful.

* Kurt Cagle: Please note the similarity of Tony's slide with biological cells. Hmmm ...

* Cedric Berger: Has anyone asked AI to generate an image of a factual made of network base units?

* Michael Robbins: Agreed, Tony. And we’ve talked about this on LinkedIn. A constellation of domain-specific and ecosystem-based Community Knowledge Graphs and Language Models. #CKGs and #CLMs
** [[MikeBennett|Mike Bennett]]: TLO as mitochondria?
** Michael Robbins: Language is inseparable from culture and context

* Harvey King: Do KG's take a different nature when dealing with math?
** [[AndreaWesterinen|Andrea Westerinen]]: Not different, but with more rules?
** [[MichaelDeBellis|Michael DeBellis]]: Yes. An ontology uses explicit models like linear algebra. LLMs use linear algebra but computes answers based on examples, doesn't have a theoretical model of math (or other domains) as an ontology does.

* [[SusanneVejdemo|Sus Vejdemo]]: Thanks for a great talk! I love the conceptualization of embeddings as ontologies.

* [[MikeBennett|Mike Bennett]]: If you reify every edge you can give each one an analog value

* [[MarkUnderwood|Mark Underwood]]: For those interested in practical cybersec use cases (lots of chatty network data, some NLP), lot of narrow domain-specific emergent ontologies; e.g., Lambda / microservice mesh etc. mark.underwood@syf.com

* [[GaryBergCross|Gary Berg-Cross]]: I think the working memory graph is a useful early concept bring the LLMs and ontologies together but it is not so easy to capture what is the context for knowledge in this representation. I would guess this is a sub-set of the fluid knowledge of what human cognition employs. Much remains unconscious. But with research AI systems may make more of this explicit.

* [[RaviSharma|Ravi Sharma]]: My query is what is the relationship among reification, provenance and context history?

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Tony, “LLMs for compute” and “As much data into graph, then translating the graph paths to NL and adding to the LLM”

* [[GaryBergCross|Gary Berg-Cross]]: Reality may be atomistically discrete but at such a nano-level that continuous models make better predictions than discrete models that are orders of magnitude too gross rather than fine grained.

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Kurt: “Community Language Models - decentralized, federated, ad hoc network of information”

* [[GaryBergCross|Gary Berg-Cross]]: Models need to be more than federated. Because we center on semantic accuracy and relevance they need to be semantically harmonized.

== Resources ==
* [https://bit.ly/3S040lR Kurt Cagle Slides]
* [https://bit.ly/46A3EH2 Tony Seale Slides]
* [https://bit.ly/46DJyvo Video Recording]
* [https://youtu.be/TAjkzg_lhVo YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 08

2023-11-12T23:23:17Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Broader thoughts]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::8 Nov 2023 17:00 GMT]]
|-
| 9:00am PST/12:00pm EST
|-
| 5:00pm GMT/6:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', strategist and blogger at [http://ailev.lievjournal.ru Laboratory Log]
** ''Knowledge graphs and large language models in cognitive architectures''
** This talk discusses styles of definitions for knowledge graphs (KG) combined with large language models (LLMs). The KG architectures and systems are reviewed, taken from Ontolog Forum's 2020 Communique. A framework is proposed for a cognitive architecture using both LLMa and KGs for the evolution of knowledge during 4E (embodied, extended, embedded, enacted) cognition. In this framework, ontologies are understood as answers to the question "What is in the world?" and can be found in representations that vary across a spectrum of formality/rigor. An example is given of the use of ontology engineering training in management, where upper-level ontologies are given to students in the form of informal course texts (with the goal of obtaining a fine-tuned LLM within the "neural networks" of students' brains) coupled with lower-level ontologies that are more formal (such as data schemas for databases and knowledge graphs).
** Anatoly Levenchuk has worked as a strategy consultant for more than 30 years. He helps with vision and strategy definition to many government agencies and large companies. Now he is science head of Aisystant that serves as a school in engineering and management. His first machine learning project was in 1977, first ontology engineering project was in 1980. He is author of several textbooks on systems thinking, methodology, systems engineering, systems management, natural and artificial intelligence, education as "person engineering". His blog "Laboratory Log" http://ailev.lievjournal.ru in Russian has more than 3,000 subscribers.
** [https://bit.ly/3sljmXt Slides]
* '''[[ArunMajumdar|Arun Majumdar]]''' and '''[[JohnSowa|John Sowa]]''', [https://permion.ai/ Permion AI]
** ''Trustworthy Computation: Diagrammatic Reasoning With and About LLMs''
** Large Language Models (LLMs) were designed for machine translation (MT). Although LLM methods cannot do any reasoning by themselves, they can often find and apply reasoning patterns that they find in the vast resources of the WWW. For common problems, they frequently find a correct solution. For more complex problems, they may construct a solution that is partially correct for some applications, but disastrously wrong or even hallucinogenic for others. Systems developed by Permion use LLMs for what they do best. But they combine them with precise and trusted methods of diagrammatic reasoning based on conceptual graphs (CGs). They take advantage of the full range of technology developed by 60+ years of AI, computer science, and computational linguistics. For any application, Permion methods derive an ontology tailored to the policies, rules, and specifications of the project or business. All programs and results they produce are guaranteed to be consistent with that ontology.
** [https://bit.ly/464bRlF John's Slides]
** [https://bit.ly/475rpH1 Arun's Slides]
* [https://bit.ly/3QMfqZj Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 8 November 2023'''
* Start Time: 9:00am PST / 12:00pm EST / 6:00pm CET / 5:00pm GMT / 1700 UTC
** Note that Daylight Saving Time has ended in Europe, Canada and the US.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=8&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[JohnSowa|John Sowa]]
* [[ArunMajumdar|Arun Majumdar]]
* [[AnatolyLevenchuk|Anatoly Levenchuk]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* [[AmitSheth|Amit Sheth]]
* [[DavidEddy|David Eddy]]
* [[DaleFitch|Dale Fitch]]
* Nancy Wiegand
* [[PeteRivett|Pete Rivett]]
* [[VictorAgroskin|Victor Agroskin]]
* MZeng
* Tony Cohn (Leeds)

Complete list of participants was not captured

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: When I think about models, it is not necessarily graph or even video but something like a kind of vision mind-based understanding and yes I express it often in language or express math using the language also.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Yes, any form. But if it (model) has patterns then you can consider it as text -- semiotics tell us "all is texts of patterns as letters". Pattern languages are about behavior as a text (chain of patterns). Mathematics is about patterns too.

* [[MichaelDeBellis|Michael DeBellis]]: I don't understand what the speaker means by quantum/digital memory.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: It is about exact copy of information, information is about difference. 1 bit is a result of measurement. To evolution can proceed, you need genes as exactly copied information about results of previous evolution steps. If you have not digital/quantum/discrete form for exact copying, you cannot accumulate knowledge, error in analog form will be prohibitive for evolution.

* [[RaviSharma|Ravi Sharma]]: Second bullet can you expand how cognition and quantum memory are understood?
** [[MikeBennett|Mike Bennett]]: That is the 2nd bullet of Slide 3 (for later reference in the Q&A)
** [[AndreaWesterinen|Andrea Westerinen]]: This is discussed at the end of the talk, in the Q&A

* [[RaviSharma|Ravi Sharma]]: Generative and interactive models, how will these be integrated at different levels?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Generative and discriminative (not interactive) models. You can ask about any models, but get different types of answers: possible worlds descriptions from generative models and classification labels from discriminative models. With many nuances, sure )))
** [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk I think your distinction between "interactive" models is very important!!! See for example the interactionist paradigm of computing by Peter Wegner and Dinah Goldin.

* [[RaviSharma|Ravi Sharma]]: Anatoly would you call L4 Contemplative?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: No. All models about activity (enactive cognition, activity is everywhere).

* [[RaviSharma|Ravi Sharma]]: Generation Differentiation is described, how do you integrate to get all facts together that means knowledge?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: I simply deny all guesses that give problems in inference. All survived guesses are integrated (i.e., not give errors in inference).

* [[RaviSharma|Ravi Sharma]]: Last Q for Anatoly, how do you introduce value systems in these cognitive architectures other than through social media etc.

* [[DouglasMiles|Douglas Miles]]: The most interesting and confusing thing about LLMs is we have no idea how to teach them any new skills... other than: We fire a hose of data and text at them and just pray

* [[AmitSheth|Amit Sheth]]: About passing Turing Test: Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as You May Think -- Introducing AI Detectability Index, https://lnkd.in/gr6cizEZ
** [[ArunMajumdar|Arun Majumdar]]: The distinction between "human" and machine is being attacked on stylometric principles but this is very hard as computer GPT outputs are mixed with human inputs ...

* [[RaviSharma|Ravi Sharma]]: What efforts are likely to succeed in making ChatGPT more accurate?
** [[ArunMajumdar|Arun Majumdar]]: @Ravi Sharma so our approach is to use Conceptual Graphs as our formal knowledge graph approach to creating a "surrogate" model to drive the LLM/GPTs.

* [[ArunMajumdar|Arun Majumdar]]: We (John and I) are not focused on the detection problem because that is something that is not related to our primary focus in knowledge graphs (Conceptual Graphs) as a formalism for symbolic AI with Generative AI.
** [[ArunMajumdar|Arun Majumdar]]: I can share screenshots that compares what we do to what others do. Basically, zero-hallucinations. We can look at a simple hallucination problem later after John's talk if needed. Ken has my screenshots.

* [[RaviSharma|Ravi Sharma]]: Is math not derived from metaphysics?
** [[ArunMajumdar|Arun Majumdar]]: Math is one of the great abstractions of human kind that is a formal but extremely open field for creativity. The key is that creativity in math is often understated. Closed World Models (CWMs) which include LLMs are not capable of creating outside something that they have been input (in their ML training, even with human reinforcement feedback learning).
** [[ArunMajumdar|Arun Majumdar]]: Tuning a GPT or LLM is still a closed world model.
** [[ArunMajumdar|Arun Majumdar]]: Semiotic models differ - the semiotics of Saussure are different to Pierce. One is dyadic, and the other is triadic. So the semiotics matter. These distinctions are not involved in GPT/LLM constructs. However, this important distinction may be approachable in the future.
** [[ArunMajumdar|Arun Majumdar]]: The patterns in mathematics are not probabilistic. They are driven by rational principles. So these are different things.
** [[ArunMajumdar|Arun Majumdar]]: You can use probability, like you can use a million monkeys typing with probabilistic bias to get some pattern candidates but there is no deep insight or principle driving these outputs.

* [[RaviSharma|Ravi Sharma]]: LLM and cognitive scientists and cog-memory well explained.

* [[JanetSinger|Janet Singer]]: Question for Anatoly re 4E cognition. Shift to 4E seems to be the key for appreciating and taking advantage of human-machine teaming opportunities post LLM revolution. But do you see 4E as applicable to software AIs separated from humans? Or just to robots (arguably embodied and embedded if in limited sense)? Or is it that we are the 4E extensions of the software AIs?

* [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk Behavior is not truly just a random process. There is "intention" in living beings - such as the need to survive and thrive. Stochastics are a way to study behavior, to mimic some behaviors but the complexity of goal-driven and intentful behavior is actually much more complex than probabilities.

* [[ArunMajumdar|Arun Majumdar]]: Permion uses tensor mathematics but in the logical formalisms so that predications and symbolic logic has a mapping to/from from the tensorial structures. There are several emerging developments including, for example, the exploration of alternatives algebras such as Clifford or Geometric algebras beyond the conventional real valued Gibbs algebras used nowadays.

* [[DavidEddy|David Eddy]]: 1.5M LoC is NOT a big system

* [[DouglasMiles|Douglas Miles]]: The earth may not be round enough to have a completely circular cord
** [[DouglasMiles|Douglas Miles]]: It takes a person that had experience beyond words to answer that
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a thought experiment discussed by John Sowa - about a circular cord around the earth's equator, and adding 1 yard to it. It raises the cord to approx 6" above the ground.

* [[MikeBennett|Mike Bennett]]: For those of us not from the US, what is the capital of Alaska? It feels like this is a vital plot point.
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a ChatGPT response in John's/Arun's presentation - that Juneau is the capital of Alaska, but there have been attempts to move the capital to another city. And, the underlying issue is whether this information was in the provided training data/corpus (e.g., could you track provenance?)

* [[RaviSharma|Ravi Sharma]]: Arun Say more about scaffolding. Is it only E and R and detailed parsed later, but using what rules?
** [[ArunMajumdar|Arun Majumdar]]: Scaffolding is based on purely mathematical methods based on the Zipf distribution laws of terminologies and the H-Point of the Zipf distribution and Eigen computation methods to identify the modules of graph structures from the text. Scaffolding provides a language
agnostic foundation.

* [[DaleFitch|Dale Fitch]]: Is your "logic" an ontology?
** [[ArunMajumdar|Arun Majumdar]]: Yes we induce and also human update the "ontology" but the logic is using the ontology

* [[DavidEddy|David Eddy]]: Thank you Ken, John, Arun & Anatoly.

== Resources ==
* [https://bit.ly/3QMfqZj Video Recording]
* [https://youtu.be/G4S-Zrc5qUk YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 08

2023-11-12T23:20:12Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Broader thoughts]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::8 Nov 2023 17:00 GMT]]
|-
| 9:00am PST/12:00pm EST
|-
| 5:00pm GMT/6:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', strategist and blogger at [http://ailev.lievjournal.ru Laboratory Log]
** ''Knowledge graphs and large language models in cognitive architectures''
** This talk discusses styles of definitions for knowledge graphs (KG) combined with large language models (LLMs). The KG architectures and systems are reviewed, taken from Ontolog Forum's 2020 Communique. A framework is proposed for a cognitive architecture using both LLMa and KGs for the evolution of knowledge during 4E (embodied, extended, embedded, enacted) cognition. In this framework, ontologies are understood as answers to the question "What is in the world?" and can be found in representations that vary across a spectrum of formality/rigor. An example is given of the use of ontology engineering training in management, where upper-level ontologies are given to students in the form of informal course texts (with the goal of obtaining a fine-tuned LLM within the "neural networks" of students' brains) coupled with lower-level ontologies that are more formal (such as data schemas for databases and knowledge graphs).
** Anatoly Levenchuk has worked as a strategy consultant for more than 30 years. He helps with vision and strategy definition to many government agencies and large companies. Now he is science head of Aisystant that serves as a school in engineering and management. His first machine learning project was in 1977, first ontology engineering project was in 1980. He is author of several textbooks on systems thinking, methodology, systems engineering, systems management, natural and artificial intelligence, education as "person engineering". His blog "Laboratory Log" http://ailev.lievjournal.ru in Russian has more than 3,000 subscribers.
** [https://bit.ly/3sljmXt Slides]
* '''[[ArunMajumdar|Arun Majumdar]]''' and '''[[JohnSowa|John Sowa]]''', [https://permion.ai/ Permion AI]
** ''Trustworthy Computation: Diagrammatic Reasoning With and About LLMs''
** Large Language Models (LLMs) were designed for machine translation (MT). Although LLM methods cannot do any reasoning by themselves, they can often find and apply reasoning patterns that they find in the vast resources of the WWW. For common problems, they frequently find a correct solution. For more complex problems, they may construct a solution that is partially correct for some applications, but disastrously wrong or even hallucinogenic for others. Systems developed by Permion use LLMs for what they do best. But they combine them with precise and trusted methods of diagrammatic reasoning based on conceptual graphs (CGs). They take advantage of the full range of technology developed by 60+ years of AI, computer science, and computational linguistics. For any application, Permion methods derive an ontology tailored to the policies, rules, and specifications of the project or business. All programs and results they produce are guaranteed to be consistent with that ontology.
** [https://bit.ly/464bRlF John's Slides]
** [https://bit.ly/475rpH1 Arun's Slides]
* [https://bit.ly/3QMfqZj Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 8 November 2023'''
* Start Time: 9:00am PST / 12:00pm EST / 6:00pm CET / 5:00pm GMT / 1700 UTC
** Note that Daylight Saving Time has ended in Europe, Canada and the US.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=8&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[JohnSowa|John Sowa]]
* [[ArunMajumdar|Arun Majumdar]]
* [[AnatolyLevenchuk|Anatoly Levenchuk]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* [[AmitSheth|Amit Sheth]]
* [[DavidEddy|David Eddy]]
* [[DaleFitch|Dale Fitch]]

Complete list of participants was not captured

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: When I think about models, it is not necessarily graph or even video but something like a kind of vision mind-based understanding and yes I express it often in language or express math using the language also.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Yes, any form. But if it (model) has patterns then you can consider it as text -- semiotics tell us "all is texts of patterns as letters". Pattern languages are about behavior as a text (chain of patterns). Mathematics is about patterns too.

* [[MichaelDeBellis|Michael DeBellis]]: I don't understand what the speaker means by quantum/digital memory.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: It is about exact copy of information, information is about difference. 1 bit is a result of measurement. To evolution can proceed, you need genes as exactly copied information about results of previous evolution steps. If you have not digital/quantum/discrete form for exact copying, you cannot accumulate knowledge, error in analog form will be prohibitive for evolution.

* [[RaviSharma|Ravi Sharma]]: Second bullet can you expand how cognition and quantum memory are understood?
** [[MikeBennett|Mike Bennett]]: That is the 2nd bullet of Slide 3 (for later reference in the Q&A)
** [[AndreaWesterinen|Andrea Westerinen]]: This is discussed at the end of the talk, in the Q&A

* [[RaviSharma|Ravi Sharma]]: Generative and interactive models, how will these be integrated at different levels?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Generative and discriminative (not interactive) models. You can ask about any models, but get different types of answers: possible worlds descriptions from generative models and classification labels from discriminative models. With many nuances, sure )))
** [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk I think your distinction between "interactive" models is very important!!! See for example the interactionist paradigm of computing by Peter Wegner and Dinah Goldin.

* [[RaviSharma|Ravi Sharma]]: Anatoly would you call L4 Contemplative?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: No. All models about activity (enactive cognition, activity is everywhere).

* [[RaviSharma|Ravi Sharma]]: Generation Differentiation is described, how do you integrate to get all facts together that means knowledge?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: I simply deny all guesses that give problems in inference. All survived guesses are integrated (i.e., not give errors in inference).

* [[RaviSharma|Ravi Sharma]]: Last Q for Anatoly, how do you introduce value systems in these cognitive architectures other than through social media etc.

* [[DouglasMiles|Douglas Miles]]: The most interesting and confusing thing about LLMs is we have no idea how to teach them any new skills... other than: We fire a hose of data and text at them and just pray

* [[AmitSheth|Amit Sheth]]: About passing Turing Test: Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as You May Think -- Introducing AI Detectability Index, https://lnkd.in/gr6cizEZ
** [[ArunMajumdar|Arun Majumdar]]: The distinction between "human" and machine is being attacked on stylometric principles but this is very hard as computer GPT outputs are mixed with human inputs ...

* [[RaviSharma|Ravi Sharma]]: What efforts are likely to succeed in making ChatGPT more accurate?
** [[ArunMajumdar|Arun Majumdar]]: @Ravi Sharma so our approach is to use Conceptual Graphs as our formal knowledge graph approach to creating a "surrogate" model to drive the LLM/GPTs.

* [[ArunMajumdar|Arun Majumdar]]: We (John and I) are not focused on the detection problem because that is something that is not related to our primary focus in knowledge graphs (Conceptual Graphs) as a formalism for symbolic AI with Generative AI.
** [[ArunMajumdar|Arun Majumdar]]: I can share screenshots that compares what we do to what others do. Basically, zero-hallucinations. We can look at a simple hallucination problem later after John's talk if needed. Ken has my screenshots.

* [[RaviSharma|Ravi Sharma]]: Is math not derived from metaphysics?
** [[ArunMajumdar|Arun Majumdar]]: Math is one of the great abstractions of human kind that is a formal but extremely open field for creativity. The key is that creativity in math is often understated. Closed World Models (CWMs) which include LLMs are not capable of creating outside something that they have been input (in their ML training, even with human reinforcement feedback learning).
** [[ArunMajumdar|Arun Majumdar]]: Tuning a GPT or LLM is still a closed world model.
** [[ArunMajumdar|Arun Majumdar]]: Semiotic models differ - the semiotics of Saussure are different to Pierce. One is dyadic, and the other is triadic. So the semiotics matter. These distinctions are not involved in GPT/LLM constructs. However, this important distinction may be approachable in the future.
** [[ArunMajumdar|Arun Majumdar]]: The patterns in mathematics are not probabilistic. They are driven by rational principles. So these are different things.
** [[ArunMajumdar|Arun Majumdar]]: You can use probability, like you can use a million monkeys typing with probabilistic bias to get some pattern candidates but there is no deep insight or principle driving these outputs.

* [[RaviSharma|Ravi Sharma]]: LLM and cognitive scientists and cog-memory well explained.

* [[JanetSinger|Janet Singer]]: Question for Anatoly re 4E cognition. Shift to 4E seems to be the key for appreciating and taking advantage of human-machine teaming opportunities post LLM revolution. But do you see 4E as applicable to software AIs separated from humans? Or just to robots (arguably embodied and embedded if in limited sense)? Or is it that we are the 4E extensions of the software AIs?

* [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk Behavior is not truly just a random process. There is "intention" in living beings - such as the need to survive and thrive. Stochastics are a way to study behavior, to mimic some behaviors but the complexity of goal-driven and intentful behavior is actually much more complex than probabilities.

* [[ArunMajumdar|Arun Majumdar]]: Permion uses tensor mathematics but in the logical formalisms so that predications and symbolic logic has a mapping to/from from the tensorial structures. There are several emerging developments including, for example, the exploration of alternatives algebras such as Clifford or Geometric algebras beyond the conventional real valued Gibbs algebras used nowadays.

* [[DavidEddy|David Eddy]]: 1.5M LoC is NOT a big system

* [[DouglasMiles|Douglas Miles]]: The earth may not be round enough to have a completely circular cord
** [[DouglasMiles|Douglas Miles]]: It takes a person that had experience beyond words to answer that
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a thought experiment discussed by John Sowa - about a circular cord around the earth's equator, and adding 1 yard to it. It raises the cord to approx 6" above the ground.

* [[MikeBennett|Mike Bennett]]: For those of us not from the US, what is the capital of Alaska? It feels like this is a vital plot point.
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a ChatGPT response in John's/Arun's presentation - that Juneau is the capital of Alaska, but there have been attempts to move the capital to another city. And, the underlying issue is whether this information was in the provided training data/corpus (e.g., could you track provenance?)

* [[RaviSharma|Ravi Sharma]]: Arun Say more about scaffolding. Is it only E and R and detailed parsed later, but using what rules?
** [[ArunMajumdar|Arun Majumdar]]: Scaffolding is based on purely mathematical methods based on the Zipf distribution laws of terminologies and the H-Point of the Zipf distribution and Eigen computation methods to identify the modules of graph structures from the text. Scaffolding provides a language
agnostic foundation.

* [[DaleFitch|Dale Fitch]]: Is your "logic" an ontology?
** [[ArunMajumdar|Arun Majumdar]]: Yes we induce and also human update the "ontology" but the logic is using the ontology

* [[DavidEddy|David Eddy]]: Thank you Ken, John, Arun & Anatoly.

== Resources ==
* [https://bit.ly/3QMfqZj Video Recording]
* [https://youtu.be/G4S-Zrc5qUk YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 01

2023-11-11T23:57:48Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Demos of information extraction via hybrid systems]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::1 Nov
2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* [[AndreaWesterinen|Andrea Westerinen]], Creator of [https://github.com/ontoinsights/deep_narrative_analysis DNA, Deep Narrative Analysis]
** '''Title:''' Populating Knowledge Graphs: The Confluence of Ontology and Large Language Models
** '''Abstract:''' Ontology-based Knowledge Graphs (KGs) stand at the forefront of semantic data representation, providing structured views of the data in complex domains. Traditionally, populating these KGs from unstructured text involved convoluted natural language analyses and custom code, but the environment has changed with the use of Large Language Models (LLMs). This talk explores one use case - the population of a KG from news articles. The evolution of the DNA application from employing spaCy APIs to OpenAI is described, and the current (open-source) implementation discussed. Implementation issues such as sourcing the data, LLM prompts, mapping the LLM responses onto the ontology, and populating the knowledge graph are overviewed.
** [https://bit.ly/49tTuJY Slides]
* [[PrasadYalamanchi|Prasad Yalamanchi]], [https://leadsemantics.com/ Lead Semantics] CTO
** '''Title:''' Harvest Knowledge From Language - Harness the power of Large Language Models and Semantic Technology
** '''Abstract:''' Language (both text and voice) holds much of the accessible knowledge to humans. It is also the best store of the collective human knowledge. Historically, accessing this knowledge, was manual and up until recent times has progressed to varying degrees of semi-automated methods! But, with the advent of Language Models and particularly Large Language Models in the last couple of years, a fully automated access to knowledge carried in language is now becoming a reality! TextDistil, the software product from Lead Semantics applies LLMs and Ontologies to extract computable knowledge in the form of RDF triples from Text.
** [https://bit.ly/3tY3niI Slides]
* [https://bit.ly/45Z1fUY Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 1 November 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 5:00pm CET / 4:00pm GMT / 1600 UTC
** Note that Daylight Saving Time has ended in Europe but not in the US or Canada.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=1&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[PrasadYalamanchi|Prasad Yalamanchi]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[JohnSowa|John Sowa]]
* [[BartGajderowicz|Bart Gajderowicz]]
* [[GaryBergCross|Gary Berg-Cross]]
* Zefi Kavvadia
* ariusz (Telicent)
* Dan (Telicent)
* [[MarkUnderwood|Mark Underwood]]
* Sundos Al Subhi
* Jeff
* James Logan

Complete list of participants was not captured

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: What is the normal accuracy and does the accuracy of triples vary by language?
** [[RaviSharma|Ravi Sharma]]: Namely would it just depend on the language only or on domain concept would affect the KG?

* [[MichaelDeBellis|Michael DeBellis]]: Is the ontology COMPLETELY created from the Corpus, or do you start from a foundation ontology and extend it based on the Corpus?

* [[RaviSharma|Ravi Sharma]]: You implied preprocessing and semantic understanding by humans before the KG is generated? how much effort is it?

* [[RaviSharma|Ravi Sharma]]: Is there a possibility to reduce the duration by compromising the accuracy somewhat?

* [[SusanneVejdemo|Susanne Vejdemo]]: I understand we have a KG constructed from the unstructured docs. And then there’s translation of your query to triples? I am a bit uncertain where the LLM comes into this?
** [[JanetSinger|Janet Singer]]: My question as well — how exactly does the LLM come in?
** [[AndreaWesterinen|Andrea Westerinen]]: I will address this in my presentation, but can’t answer for Prasad.
** [[PrasadYalamanchi|Prasad Yalamanchi]]: LLM is coming in multiple places in the TextDistil pipeline. Once at the final summary string of the result items. It comes in building the KG, as well
** [[BartGajderowicz|Bart Gajderowicz]]: Prasad, what does the LLM do in building the KG?

* [[RaviSharma|Ravi Sharma]]: Can it show the visuals during progress such as the KG?

* [[DouglasMiles|Douglas Miles]]: Chatgpt-3.5 in some instances can work well enough to be used ovber ChatGPT-4 ?
** [[AndreaWesterinen|Andrea Westerinen]]: It MAY, but I have found profound differences. Linguistic analysis is much better in 4
** [[DouglasMiles|Douglas Miles]]: ChatGPT-3.5 can be so much faster with its return results.. I've considered running both to see when 3.5 was sufficient.. admnittely mostly it isn't.. but "convert this to owl" often is "acceptable"
** [[AndreaWesterinen|Andrea Westerinen]]: @Douglas Miles Not sure that I agree about acceptability.
** [[DouglasMiles|Douglas Miles]]: ok true.. I mean 3.5 cant even begin to convert to CLIF of CycL .. whereas it at least tried with RDF/OWL

* [[RaviSharma|Ravi Sharma]]: If there are no human interventions, how much is the KG affected?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi, that's my question as well. IMO it is usually better to have a person in the loop because creating a well designed ontology completely from a Corpus seems like the resulting ontology may not be well designed. That's why I asked the question about starting from a basic ontology and then extending that ontology, rather than creating the entire ontology from scratch

* [[DouglasMiles|Douglas Miles]]: Prasad presentation was very awesome!

* [[RaviSharma|Ravi Sharma]]: Prasad - If you do two such exercises, is the result the same/repeatable?
** [[PrasadYalamanchi|Prasad Yalamanchi]]: There is a language interpreatation to map the query string to the Ontology.
** [[PrasadYalamanchi|Prasad Yalamanchi]]: SO, if two queries (exercises as you mentioned) result in the same interpretation, then the final answers will be the same

* [[DouglasMiles|Douglas Miles]]: Something that impresses me and is unique about Andrea's work (even year or two ago.. ) ... She actually supports full modality representations in RDF-ish languages.. Stuff that normally I would only dare to use CLIF to represent!

* [[RaviSharma|Ravi Sharma]]: Are there similarities to the rhetoric possibilities of metaphor, context, explanation, etc to improve your results?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure what is being asked. I am exposing the use of rhetorical devices to help readers understand how the text might be affecting their interpretations of it.

* [[RaviSharma|Ravi Sharma]]: Andrea does ML or AI enter this exercise? And results you showed, if so where?
* [[RaviSharma|Ravi Sharma]]: I mean what ML and learning sets were used in OpenAI
** [[AndreaWesterinen|Andrea Westerinen]]: OpenAI's complete technology stack is not disclosed but their website says "We build our generative models using a technology called deep learning, which leverages large amounts of data to train an AI system to perform a task."

* [[DouglasMiles|Douglas Miles]]: To help answer how LLMs can be useful in translation: https://chat.openai.com/share/039d72c3-8432-48d1-98b8-63e15614bbef

* [[MarkUnderwood|Mark Underwood]]: Excellent presentations and important work for the ontology community

* Sundos Al Subhi: Thank you all!! Great information.

* [[JanetSinger|Janet Singer]]: Excellent presentations — Looking forward to seeing these ideas integrated in the future session(s)

* [[DouglasMiles|Douglas Miles]]: (there is no question that the KR Andrea is doing is rock solid!) Here is my question though: Are any of the RDF reasoners good enough to do the reasoning/query that Andrea expects?
** [[AndreaWesterinen|Andrea Westerinen]]: @Douglas Miles Yes, I use Stardog. Also allows use of Voicebox which encode NL queries in SPARQL!

* [[DouglasMiles|Douglas Miles]]: Thank you that was great!

* Dan (Telicent): Thank you for your presentations 🙂

* Zefi Kavvadia : thank you!

* Mariusz (Telicent) : Thank you.

== Resources ==
* [https://bit.ly/45Z1fUY Video Recording]
* [https://youtu.be/h9Rl6nZ50Ls YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_01]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_01]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 04

2023-11-11T23:03:16Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Overview]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::4 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Conveners
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==

'''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''

'''Title:''' ''Fall Series Kickoff and Overview''

'''Abstract:''' The opening session of the Ontology Summit 2024 Fall Series overviews the LLM, ontology and knowledge graph landscapes, as well as introducing the participating speakers. The goal of the Series is to understand, discuss and debate the similarities, differences and overlaps across these landscapes. In addition, we will use these sessions to help to formulate the full 2024 Summit.

[https://bit.ly/3Q28U00 Slides]

[https://bit.ly/3rCnyC0 Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 4 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=04&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[AlexShkotin|Alex Shkotin]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[ToddSchneider|Todd Schneider]]
* [[MikeBennett|Mike Bennett]]
* Ayya Niyyanika Bhikkhuni
* Bill McCarthy
* Zefi Kavvadia
* [[RamSriram|Ram D Sriram]]
* Andrew McCaffrey
* Steve Wartik
* [[BartGajderowicz|Bart Gajderowicz]]
* [[MarkFox|Mark Fox]]
* Seungmin Seo
* JL Valente
* [[MichaelDeBellis|Michael DeBellis]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* Sima Yazdani
* [[JohnSowa|John Sowa]]
* [[KenBaclawski|Ken Baclawski]]
* [[RaviSharma|Ravi Sharma]]
* Sergey Rodionov
* Taj Uddin
* [[MarkRessler|Mark Ressler]]
* Asiyah Yu Lin
* Hayden Spence
* Michael Singer
* Roberta Ferrario
* Chris Novell
* Emanuele Bottazzi
* Marco Monti
* [[JanetSinger|Janet Singer]]

== Discussion ==
* [[MikeBennett|Mike Bennett]]: Andrea's quote: "Ontologies are the backing definitions behind knowledge graphs" is a great way of describing the distinction between them.
** Emanuele Bottazzi: Or justifications

* [[ToddSchneider|Todd Schneider]]: Many Knowledge Graphs are not based on an ontology.
** [[AlexShkotin|Alex Shkotin]]: but keep it inside
** [[MikeBennett|Mike Bennett]]: I would not characterize such a thing as a knowledge graph, even if it re-uses that label for itself. Whence the claim of 'Knowledge' in KG if not semantics? Might not be an OWL-ology of course.
** [[BartGajderowicz|Bart Gajderowicz]]: An ontology is the “schema” for a knowledge graph, so it may not be designed well but there is a “schema” that defines nodes and edges in some way.

* Steven Wartik: I like to distinguish between a KG and a knowledge base. A KG is a graph. It doesn't necessarily have a schema. A KB is a KG whose schema is an ontology. This is just terminology, but I find it helps my sponsors understand.

* [[AlexShkotin|Alex Shkotin]]: Give me KG and I'll extract it's ontology.

* [[KenBaclawski|Ken Baclawski]]: KGs were covered in Ontology Summit 2020. The communique has precise definitions: https://ontologforum.s3.amazonaws.com/OntologySummit2020/Communique/OntologySummit2020Communique.pdf

* [[ToddSchneider|Todd Schneider]]: Knowledge =def. “facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject” (from New Oxford American Dictionary)

* [[ToddSchneider|Todd Schneider]]: ‘Meaning’ is an ambiguous term.
** Andrew McCaffrey: To "table" a motion means completely the opposite things in the US and the UK. :D

* Ayya Niyyanika Bhikkhuni: As generative AI hallucinations become an issue, there seems a need for credibility scoring. I am about a decade out-of-the-loop, but know we were talking about this many summits ago. This is in regards to trust.
** [[BartGajderowicz|Bart Gajderowicz]]: Explanations WITH hallucinations are a huge problem for LLMs. They sound credible, and may be logically sound, but are completely wrong.
** Emanuele Bottazzi: Perhaps all the probabilistic approaches cannot be explanatory, since they “happen” to be wrong or right
** [[BartGajderowicz|Bart Gajderowicz]]: Ideally the explanation would come from explicit knowledge. Most LLMs just don’t have that. Ensemble ML architectures may include explicit knowledge somewhere, but if the underlying processes and representations are probabilistic we reach a hard limit on explainability. Of course you can have an explanation that provides “certainty” about the answer and explanation, which is often sufficient.
** Emanuele Bottazzi: I would add that ideally the explanation would come from the explicit _use_ of knowledge and principles

* [[BartGajderowicz|Bart Gajderowicz]]: Do LLMs perform natural language understanding (NLU), or just processing (NLP)?
** [[BartGajderowicz|Bart Gajderowicz]]: Given my definition of knowledge I’d say NLP only. Even a simple Word2vec embedding is able to identify similarity between complex objects, but I would not consider it understanding (or knowledge)

* Ayya Niyyanika Bhikkhuni: “What is really true” is the underlying question when translating ancient text. The project I am working on is taking translations from humans and Generative AI and it is hoped then that people practicing according to their interpretation of the texts would tune the translations based on ‘tacit knowledge.’

* Marco Monti: QUESTION: if neither LLM models nor Knowledge Graphs allow for compositionality and high contextualization of answers from a chat bot, what are the mechanisms behind the scenes of GPT X to answer so punctually and contextually ?

* [[JanetSinger|Janet Singer]]: Yes, mimicry is the key characterization of what LLMs do. Parallels the 1950s it was thought that mimicry of biological behavior would inevitably lead to a structural model of living systems, and then to artificially generated life itself. See critiques by Robert Rosen.

* [[GaryBergCross|Gary Berg-Cross]]: LLM based systems can learn on the job although you wouldn't call it based on experience. This has been said about the learning: "When a user interacts with an LLM-based system, the system is able to observe the user's responses and learn from them. This allows the system to improve its ability to generate responses that are relevant to the user's needs.

* [[GaryBergCross|Gary Berg-Cross]]: There are a number of ways that LLM-based systems can be trained using chat responses. One common approach is to use reinforcement learning. In reinforcement learning, the system is rewarded for generating responses that are positive and helpful. This encourages the system to learn what kinds of responses are most likely to be well-received by users."

* [[ToddSchneider|Todd Schneider]]: Could explain “links in OWL are not first class objects”?
** Steven Wartik: Todd, a first-class object is uniquely identifiable. A reified triple is a 1st-class object.
** Asiyah Yu Lin: I think the knowledge graph users who doesn't care too much about OWL thinking of data level or instance level. The ontology is really about classes. There is a blurred line between what is data and what is class.
** [[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Suppose you have a model of a highway as a graph where nodes are cities and links are roads. You want to model the time it takes to get from two nodes as information directly on the link. You can do that with Neo4J but now with OWL. With OWL you need to use the design pattern where you reify the relation with a new class.
** [[ToddSchneider|Todd Schneider]]: Michael, thank you for the explanation. Per your example, it could be the case that the representation (of the entities and their relations) was inadequate to support the query (i.e. with reification). Typo “ with reification’ should be ‘Without reification).
** [[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Yes. My question is how easy is it to take an OWL ontology where you have reified the relations and use graph theoretic algorithms? I don't know because I haven't used these algorithms in a long time. One thing I'm thinking about is creating an extension to OWL (I mean things like new classes and Python or SPARQL) where when you assert a new property value you have the option to create an instance of a Relation class and store data directly on that instance. That way you could treat the OWL ontology as a true graph.
** [[MichaelDeBellis|Michael DeBellis]]: Often you can even ask GPT-4 to create the KIF or CycL or CLIF .. and it will

* Hayden Spence: RE: Generating ontologies with LLMs: https://github.com/monarch-initiative/ontogpt

* [[JanetSinger|Janet Singer]]: Here, mimicry of knowledge-driven behavior is being promoted as inevitably leading to structural models of knowledge and then to ‘emergent consciousness’. Ontologies are structural (good for modeling within their scope); LLMs are behavioral
** [[JanetSinger|Janet Singer]]: Here as in the hype cycle, not by Andrea 🙂

* [[Douglas Miles|Douglas Miles]]: GPT-3 btw seems useless compared to GPT-4 on this front

* Hayden Spence: From my understanding, GPT-4 is multimodal and multimodel in the sense its training is higher parameter, it incorporates more than just text data, and the actual interface is the interaction of multiple GPT models working together.

* [[ToddSchneider|Todd Schneider]]: What is ‘semantic understanding’?
** [[AndreaWesterinen|Andrea Westerinen]]: It is "natural language understanding"

* Hayden Spence: Is the use of established controlled vocabularies that are under license like SNOMED CT, MedDRA, ICD10/0, or standards like FHIR, and the mappings between them -- once embedded -- still restricted? At what point does transformation of information collection become its own separate from the digested information.

* [[Douglas Miles|Douglas Miles]]: i don't have a question at this point.. but love this talk!

* [[JanetSinger|Janet Singer]]: Symbolic and connectionist theories of cognition are both computationalist. Leaves out 4-E embodied cognition perspective

== Resources ==
* [https://bit.ly/3rCnyC0 Video Recording]
* [https://youtu.be/5uL5HXD4f3w YouTube Video]

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_04]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 11

2023-11-11T23:00:55Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Setting the stage]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::11 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
'''[[DeborahMcGuinness|Deborah L. McGuinness]]''' 
Rennselaer Tetherless World Senior Constellation Chair 
Professor of Computer Science, Cognitive Science, and Industrial and Systems Engineering 

'''Title''': ''The Evolving Landscape: Generative AI, Ontologies, and Knowledge Graphs''

'''Abstract''': AI is in the news with astonishing regularity and the variety of announcements is often dizzying. In this talk, we will explore some opportunities (as well as threats) from the world of generative AI with respect to semantic technologies. We will explore some questions worth pondering as we plan our ontology and knowledge graph directions and hopefully leave with some mutually beneficial synergies between large language models and classical Ontology Summit topics.

[https://bit.ly/3LYsfgh Slides]

[https://bit.ly/48PYojR Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 11 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=11&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[AndreaWesterinen|Andrea Westerinen]]
* Ayya Niyyanika Bhikkhuni
* [[BartGajderowicz|Bart Gajderowicz]]
* [[DeborahMcGuinness|Deborah McGuinness]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* John O'Gorman
* [[JohnSowa|John Sowa]]
* [[KenBaclawski|Ken Baclawski]]
* [[LeiaDickerson|Leia Dickerson]]
* [[MarkUnderwood|Mark Underwood (IS Innovation)]]
* [[MikeBennett|Mike Bennett]]
* Phil Jackson
* [[RaviSharma|Ravi Sharma]]
* [[SusanneVejdemo|Sus Vejdemo]]
* [[ToddSchneider|Todd Schneider]]

== Discussion ==
* [[RaviSharma|Ravi Sharma]] : Deborah what if the training set is not representative, how do we know the errors are due to lack of a good raining set or bad algorithms such as Bayes implementation?
* [[RaviSharma|Ravi Sharma]] : Would like to know how much richness beyond ontologies can be added to KGs?
** [[ToddSchneider|Todd Schneider]] : What about the converse: How can ontologies be leveraged to improve LLMs?
** [[ToddSchneider|Todd Schneider]] : Ontologies and knowledge graphs provide explicitness, in contrast to LLMs.
* [[BartGajderowicz|Bart Gajderowicz]] : Following this year’s FOIS in Canada, I submitted a proposal for a FOIS Working Group specifically for an ontology benchmark suite, in case anyone is interested in joining me in the effort.
** [[ToddSchneider|Todd Schneider]] : Bart, ‘ontology benchmark suite’??
** [[BartGajderowicz|Bart Gajderowicz]] : Yes, it’s exactly what is it sounds like. At FOIS there were a few presentations that used datasets and a few that found issues with published ontologies. It came from these works
* [[RaviSharma|Ravi Sharma]] : Are LLM aware of graphic ways of understanding or even ability to create imagelike information from patterns?
* [[MikeBennett|Mike Bennett]] : I would expect a wine ontology to have 'Relative' concepts like terroir (land in wine context); vintage (time in wine context); varietal (grape in wine context) etc. i.e. contextually relevant concepts. The reason I mention it: how does any LLM recognize or relate to contextually relative concepts?
* [[RaviSharma|Ravi Sharma]] : Seems like Agile development type template
* [[RaviSharma|Ravi Sharma]] : In wine ontology, what would be impact of adding one or two more variables?
* [[SusanneVejdemo|Sus Vejdemo]] : I'm creating an ontology for the linguistics of temperature terms (how different language communities divide up the sensation of hot vs cold in different ways). I used chatGPT to help me with OWL syntax - but if I hadn't already had a very firm grasp of what I wanted to know, it would have led me pretty wrong.
* [[DouglasMiles|Douglas Miles]] : LLMs turn out to be the best version of a search engine at finding obscure data
** [[DouglasMiles|Douglas Miles]] : (not calling *this* obscure, but it is ideal for finding exactly the works we are looking for sometimes)
* [[SusanneVejdemo|Sus Vejdemo]] : Beware - chatGPT can sometimes delete classes without informing you if you run bigger ontologies through it.
** [[RaviSharma|Ravi Sharma]] : Sue that is why context and prior knowledge are good filters?
* [[DouglasMiles|Douglas Miles]] : often i will ask "What information was just left out"
** [[DouglasMiles|Douglas Miles]] : and the LLM usually (with enough harassment) will supply me with what i wanted
* [[MikeBennett|Mike Bennett]] : I like ChatBS as a term. LLM is technically bullshit.
* Phil Jackson : Can LLM's perform 'self-talk' yet, e.g. to emulate artificial consciousness?
** [[BartGajderowicz|Bart Gajderowicz]] : There’s a recent paper that evaluates whether the LLM knows it’s hallucinating, that may be close to its internet “thinking” as it explores different text it wants to generate: Azaria, A., & Mitchell, T. (2023). The Internal State of an LLM Knows When its Lying. http://arxiv.org/abs/2304.13734
** Phil Jackson : thanks for this reference
** [[SusanneVejdemo|Sus Vejdemo]] : We'd want it to have an "evidentiality" signal, like some natural languages do!
* [[DouglasMiles|Douglas Miles]] : Chat GPT3.5 or 4 there?
** [[DouglasMiles|Douglas Miles]] : hah forecasting
* [[MikeBennett|Mike Bennett]] : Mansplaining as a service
* [[JohnSowa|John Sowa]] : Summary of all these messages: LLMs are flaky. If you're lucky, they're great. If not, you have no idea what went wrong. That is not acceptable for any mission-critical applications.
* [[RaviSharma|Ravi Sharma]] : What would have the outcome been if you had experts key inputs to create a new exposures health ontology?
* [[GaryBergCross|Gary Berg-Cross]] : It will be good to look at these leverage and pain points again in a year
* John O'Gorman : @Semantium is using a faceted, foundational ontology
* [[DouglasMiles|Douglas Miles]] : GPT for Python code, it's kinda hit-or-miss.. mostly garbage. But with Prolog and Lisp, it's useful!.. Maybe it's because there's less bad code out there (only one place the CMU AI archive).. Tehre will be a big market for LLM curators
* [[ToddSchneider|Todd Schneider]] : We don’t even know what ‘context’ is.
** [[MikeBennett|Mike Bennett]] : Context is the nexus of time, place, role, event etc. i.e. a bunch of concepts in the ontology (or instances of these)
** John O'Gorman : @Todd Schneider - Context is the way language reduces ambiguity.
*** [[ToddSchneider|Todd Schneider]] : Yes, but that’s an application of [a] ‘context’.
** [[AndreaWesterinen|Andrea Westerinen]] : I have tried to provide ‘context’ in my prompts. For example, summarize this news article assuming it was written by a right-wing publication.
* [[DouglasMiles|Douglas Miles]] : You have to harass them over and over John .. such as "Give me three completely different translations to CLIF:
* [[AndreaWesterinen|Andrea Westerinen]] : My “sandbox” is translating NL to an ontology. Will be talking about it on Nov 1.
** [[MarkUnderwood|Mark Underwood (IS Innovation)]] : Will try to make your talk. May have possible uses for specialized, DSL-type ontologies that are constructed from emerging, fresh domains; e.g., AWS Lambda
* [[GaryBergCross|Gary Berg-Cross]] : Just as chatGPTs can do some programming they can express concepts in formal languages and thus help with Ontology population.
* [[DouglasMiles|Douglas Miles]] : "Would that translation you just gave translate back to the same English?"
* [[BartGajderowicz|Bart Gajderowicz]] : John Sowa: Speaking with those in classic literature, translation models do a poor job exactly because of the nuances in classics where meaning is lost due to the poetic styles
* [[DouglasMiles|Douglas Miles]] : I am in general agreement.. and admit there is no clear way to resolve LLM issues
* Ayya Niyyanika Bhikkhuni : The metrics are a concern even more when the “research” being done is about what something from an ancient language means and how to apply what it “means” in today’s world. So the pain point of competency metrics is not just the actual textual validity of the translation but also of “context.”
* [[SusanneVejdemo|Sus Vejdemo]] : What's a good way to get on the mailing list? I found this seminar from a Taxonomy LinkedIn group post. (I'm a linguist, semanticist)
** [[LeiaDickerson|Leia Dickerson]] : https://ontologforum.com/index.php/WikiHomePage
** [[LeiaDickerson|Leia Dickerson]] : Sign up and more info is here. Also, if you are not on it, I suggest adding yourself to the KGC Slack channel. https://www.knowledgegraph.tech/community/
** [[AndreaWesterinen|Andrea Westerinen]] : Send email to membership@ontologforum.org
* [[GaryBergCross|Gary Berg-Cross]] : I've used chats for competency Qs but not use cases which seems like an interesting possibility as part of KE sessions.

== Resources ==
* [https://bit.ly/48PYojR Video Recording]
* [https://youtu.be/BHqEjqWWVaQ YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_11]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_11]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 08

2023-11-09T17:19:31Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Broader thoughts]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::8 Nov 2023 17:00 GMT]]
|-
| 9:00am PST/12:00pm EST
|-
| 5:00pm GMT/6:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', strategist and blogger at [http://ailev.lievjournal.ru Laboratory Log]
** ''Knowledge graphs and large language models in cognitive architectures''
** This talk discusses styles of definitions for knowledge graphs (KG) combined with large language models (LLMs). The KG architectures and systems are reviewed, taken from Ontolog Forum's 2020 Communique. A framework is proposed for a cognitive architecture using both LLMa and KGs for the evolution of knowledge during 4E (embodied, extended, embedded, enacted) cognition. In this framework, ontologies are understood as answers to the question "What is in the world?" and can be found in representations that vary across a spectrum of formality/rigor. An example is given of the use of ontology engineering training in management, where upper-level ontologies are given to students in the form of informal course texts (with the goal of obtaining a fine-tuned LLM within the "neural networks" of students' brains) coupled with lower-level ontologies that are more formal (such as data schemas for databases and knowledge graphs).
** Anatoly Levenchuk has worked as a strategy consultant for more than 30 years. He helps with vision and strategy definition to many government agencies and large companies. Now he is science head of Aisystant that serves as a school in engineering and management. His first machine learning project was in 1977, first ontology engineering project was in 1980. He is author of several textbooks on systems thinking, methodology, systems engineering, systems management, natural and artificial intelligence, education as "person engineering". His blog "Laboratory Log" http://ailev.lievjournal.ru in Russian has more than 3,000 subscribers.
** [https://bit.ly/3sljmXt Slides]
* '''[[ArunMajumdar|Arun Majumdar]]''' and '''[[JohnSowa|John Sowa]]''', [https://permion.ai/ Permion AI]
** ''Trustworthy Computation: Diagrammatic Reasoning With and About LLMs''
** Large Language Models (LLMs) were designed for machine translation (MT). Although LLM methods cannot do any reasoning by themselves, they can often find and apply reasoning patterns that they find in the vast resources of the WWW. For common problems, they frequently find a correct solution. For more complex problems, they may construct a solution that is partially correct for some applications, but disastrously wrong or even hallucinogenic for others. Systems developed by Permion use LLMs for what they do best. But they combine them with precise and trusted methods of diagrammatic reasoning based on conceptual graphs (CGs). They take advantage of the full range of technology developed by 60+ years of AI, computer science, and computational linguistics. For any application, Permion methods derive an ontology tailored to the policies, rules, and specifications of the project or business. All programs and results they produce are guaranteed to be consistent with that ontology.
** [https://bit.ly/464bRlF John's Slides]
** [https://bit.ly/475rpH1 Arun's Slides]
* [https://bit.ly/3QMfqZj Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 8 November 2023'''
* Start Time: 9:00am PST / 12:00pm EST / 6:00pm CET / 5:00pm GMT / 1700 UTC
** Note that Daylight Saving Time has ended in Europe, Canada and the US.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=8&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[JohnSowa|John Sowa]]
* [[ArunMajumdar|Arun Majumdar]]
* [[AnatolyLevenchuk|Anatoly Levenchuk]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* [[AmitSheth|Amit Sheth]]
* [[DavidEddy|David Eddy]]
* [[DaleFitch|Dale Fitch]]

Complete list of participants was not captured

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: When I think about models, it is not necessarily graph or even video but something like a kind of vision mind-based understanding and yes I express it often in language or express math using the language also.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Yes, any form. But if it (model) has patterns then you can consider it as text -- semiotics tell us "all is texts of patterns as letters". Pattern languages are about behavior as a text (chain of patterns). Mathematics is about patterns too.

* [[MichaelDeBellis|Michael DeBellis]]: I don't understand what the speaker means by quantum/digital memory.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: It is about exact copy of information, information is about difference. 1 bit is a result of measurement. To evolution can proceed, you need genes as exactly copied information about results of previous evolution steps. If you have not digital/quantum/discrete form for exact copying, you cannot accumulate knowledge, error in analog form will be prohibitive for evolution.

* [[RaviSharma|Ravi Sharma]]: Second bullet can you expand how cognition and quantum memory are understood?
** [[MikeBennett|Mike Bennett]]: That is the 2nd bullet of Slide 3 (for later reference in the Q&A)
** [[AndreaWesterinen|Andrea Westerinen]]: This is discussed at the end of the talk, in the Q&A

* [[RaviSharma|Ravi Sharma]]: Generative and interactive models, how will these be integrated at different levels?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Generative and discriminative (not interactive) models. You can ask about any models, but get different types of answers: possible worlds descriptions from generative models and classification labels from discriminative models. With many nuances, sure )))
** [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk I think your distinction between "interactive" models is very important!!! See for example the interactionist paradigm of computing by Peter Wegner and Dinah Goldin.

* [[RaviSharma|Ravi Sharma]]: Anatoly would you call L4 Contemplative?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: No. All models about activity (enactive cognition, activity is everywhere).

* [[RaviSharma|Ravi Sharma]]: Generation Differentiation is described, how do you integrate to get all facts together that means knowledge?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: I simply deny all guesses that give problems in inference. All survived guesses are integrated (i.e., not give errors in inference).

* [[RaviSharma|Ravi Sharma]]: Last Q for Anatoly, how do you introduce value systems in these cognitive architectures other than through social media etc.

* [[DouglasMiles|Douglas Miles]]: The most interesting and confusing thing about LLMs is we have no idea how to teach them any new skills... other than: We fire a hose of data and text at them and just pray

* [[AmitSheth|Amit Sheth]]: About passing Turing Test: Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as You May Think -- Introducing AI Detectability Index, https://lnkd.in/gr6cizEZ
** [[ArunMajumdar|Arun Majumdar]]: The distinction between "human" and machine is being attacked on stylometric principles but this is very hard as computer GPT outputs are mixed with human inputs ...

* [[RaviSharma|Ravi Sharma]]: What efforts are likely to succeed in making ChatGPT more accurate?
** [[ArunMajumdar|Arun Majumdar]]: @Ravi Sharma so our approach is to use Conceptual Graphs as our formal knowledge graph approach to creating a "surrogate" model to drive the LLM/GPTs.

* [[ArunMajumdar|Arun Majumdar]]: We (John and I) are not focused on the detection problem because that is something that is not related to our primary focus in knowledge graphs (Conceptual Graphs) as a formalism for symbolic AI with Generative AI.
** [[ArunMajumdar|Arun Majumdar]]: I can share screenshots that compares what we do to what others do. Basically, zero-hallucinations. We can look at a simple hallucination problem later after John's talk if needed. Ken has my screenshots.

* [[RaviSharma|Ravi Sharma]]: Is math not derived from metaphysics?
** [[ArunMajumdar|Arun Majumdar]]: Math is one of the great abstractions of human kind that is a formal but extremely open field for creativity. The key is that creativity in math is often understated. Closed World Models (CWMs) which include LLMs are not capable of creating outside something that they have been input (in their ML training, even with human reinforcement feedback learning).
** [[ArunMajumdar|Arun Majumdar]]: Tuning a GPT or LLM is still a closed world model.
** [[ArunMajumdar|Arun Majumdar]]: Semiotic models differ - the semiotics of Saussure are different to Pierce. One is dyadic, and the other is triadic. So the semiotics matter. These distinctions are not involved in GPT/LLM constructs. However, this important distinction may be approachable in the future.
** [[ArunMajumdar|Arun Majumdar]]: The patterns in mathematics are not probabilistic. They are driven by rational principles. So these are different things.
** [[ArunMajumdar|Arun Majumdar]]: You can use probability, like you can use a million monkeys typing with probabilistic bias to get some pattern candidates but there is no deep insight or principle driving these outputs.

* [[RaviSharma|Ravi Sharma]]: LLM and cognitive scientists and cog-memory well explained.

* [[JanetSinger|Janet Singer]]: Question for Anatoly re 4E cognition. Shift to 4E seems to be the key for appreciating and taking advantage of human-machine teaming opportunities post LLM revolution. But do you see 4E as applicable to software AIs separated from humans? Or just to robots (arguably embodied and embedded if in limited sense)? Or is it that we are the 4E extensions of the software AIs?

* [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk Behavior is not truly just a random process. There is "intention" in living beings - such as the need to survive and thrive. Stochastics are a way to study behavior, to mimic some behaviors but the complexity of goal-driven and intentful behavior is actually much more complex than probabilities.

* [[ArunMajumdar|Arun Majumdar]]: Permion uses tensor mathematics but in the logical formalisms so that predications and symbolic logic has a mapping to/from from the tensorial structures. There are several emerging developments including, for example, the exploration of alternatives algebras such as Clifford or Geometric algebras beyond the conventional real valued Gibbs algebras used nowadays.

* [[DavidEddy|David Eddy]]: 1.5M LoC is NOT a big system

* [[DouglasMiles|Douglas Miles]]: The earth may not be round enough to have a completely circular cord
** [[DouglasMiles|Douglas Miles]]: It takes a person that had experience beyond words to answer that
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a thought experiment discussed by John Sowa - about a circular cord around the earth's equator, and adding 1 yard to it. It raises the cord to approx 6" above the ground.

* [[MikeBennett|Mike Bennett]]: For those of us not from the US, what is the capital of Alaska? It feels like this is a vital plot point.
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a ChatGPT response in John's/Arun's presentation - that Juneau is the capital of Alaska, but there have been attempts to move the capital to another city.

* [[RaviSharma|Ravi Sharma]]: Arun Say more about scaffolding. Is it only E and R and detailed parsed later, but using what rules?
** [[ArunMajumdar|Arun Majumdar]]: Scaffolding is based on purely mathematical methods based on the Zipf distribution laws of terminologies and the H-Point of the Zipf distribution and Eigen computation methods to identify the modules of graph structures from the text. Scaffolding provides a language
agnostic foundation.

* [[DaleFitch|Dale Fitch]]: Is your "logic" an ontology?
** [[ArunMajumdar|Arun Majumdar]]: Yes we induce and also human update the "ontology" but the logic is using the ontology

* [[DavidEddy|David Eddy]]: Thank you Ken, John, Arun & Anatoly.

== Resources ==
* [https://bit.ly/3QMfqZj Video Recording]
* [https://youtu.be/G4S-Zrc5qUk YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 08

2023-11-09T17:18:52Z

AndreaWesterinen:

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Broader thoughts]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::8 Nov 2023 17:00 GMT]]
|-
| 9:00am PST/12:00pm EST
|-
| 5:00pm GMT/6:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', strategist and blogger at [http://ailev.lievjournal.ru Laboratory Log]
** ''Knowledge graphs and large language models in cognitive architectures''
** This talk discusses styles of definitions for knowledge graphs (KG) combined with large language models (LLMs). The KG architectures and systems are reviewed, taken from Ontolog Forum's 2020 Communique. A framework is proposed for a cognitive architecture using both LLMa and KGs for the evolution of knowledge during 4E (embodied, extended, embedded, enacted) cognition. In this framework, ontologies are understood as answers to the question "What is in the world?" and can be found in representations that vary across a spectrum of formality/rigor. An example is given of the use of ontology engineering training in management, where upper-level ontologies are given to students in the form of informal course texts (with the goal of obtaining a fine-tuned LLM within the "neural networks" of students' brains) coupled with lower-level ontologies that are more formal (such as data schemas for databases and knowledge graphs).
** Anatoly Levenchuk has worked as a strategy consultant for more than 30 years. He helps with vision and strategy definition to many government agencies and large companies. Now he is science head of Aisystant that serves as a school in engineering and management. His first machine learning project was in 1977, first ontology engineering project was in 1980. He is author of several textbooks on systems thinking, methodology, systems engineering, systems management, natural and artificial intelligence, education as "person engineering". His blog "Laboratory Log" http://ailev.lievjournal.ru in Russian has more than 3,000 subscribers.
** [https://bit.ly/3sljmXt Slides]
* '''[[ArunMajumdar|Arun Majumdar]]''' and '''[[JohnSowa|John Sowa]]''', [https://permion.ai/ Permion AI]
** ''Trustworthy Computation: Diagrammatic Reasoning With and About LLMs''
** Large Language Models (LLMs) were designed for machine translation (MT). Although LLM methods cannot do any reasoning by themselves, they can often find and apply reasoning patterns that they find in the vast resources of the WWW. For common problems, they frequently find a correct solution. For more complex problems, they may construct a solution that is partially correct for some applications, but disastrously wrong or even hallucinogenic for others. Systems developed by Permion use LLMs for what they do best. But they combine them with precise and trusted methods of diagrammatic reasoning based on conceptual graphs (CGs). They take advantage of the full range of technology developed by 60+ years of AI, computer science, and computational linguistics. For any application, Permion methods derive an ontology tailored to the policies, rules, and specifications of the project or business. All programs and results they produce are guaranteed to be consistent with that ontology.
** [https://bit.ly/464bRlF John's Slides]
** [https://bit.ly/475rpH1 Arun's Slides]
* [https://bit.ly/3QMfqZj Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 8 November 2023'''
* Start Time: 9:00am PST / 12:00pm EST / 6:00pm CET / 5:00pm GMT / 1700 UTC
** Note that Daylight Saving Time has ended in Europe, Canada and the US.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=8&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[JohnSowa|John Sowa]]
* [[ArunMajumdar|Arun Majumdar]]
* [[AnatolyLevenchuk|Anatoly Levenchuk]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* [[AmitSheth|Amit Sheth]]
* [[DavidEddy|David Eddy]]
* [[DaleFitch|Dale Fitch]]

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: When I think about models, it is not necessarily graph or even video but something like a kind of vision mind-based understanding and yes I express it often in language or express math using the language also.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Yes, any form. But if it (model) has patterns then you can consider it as text -- semiotics tell us "all is texts of patterns as letters". Pattern languages are about behavior as a text (chain of patterns). Mathematics is about patterns too.

* [[MichaelDeBellis|Michael DeBellis]]: I don't understand what the speaker means by quantum/digital memory.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: It is about exact copy of information, information is about difference. 1 bit is a result of measurement. To evolution can proceed, you need genes as exactly copied information about results of previous evolution steps. If you have not digital/quantum/discrete form for exact copying, you cannot accumulate knowledge, error in analog form will be prohibitive for evolution.

* [[RaviSharma|Ravi Sharma]]: Second bullet can you expand how cognition and quantum memory are understood?
** [[MikeBennett|Mike Bennett]]: That is the 2nd bullet of Slide 3 (for later reference in the Q&A)
** [[AndreaWesterinen|Andrea Westerinen]]: This is discussed at the end of the talk, in the Q&A

* [[RaviSharma|Ravi Sharma]]: Generative and interactive models, how will these be integrated at different levels?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Generative and discriminative (not interactive) models. You can ask about any models, but get different types of answers: possible worlds descriptions from generative models and classification labels from discriminative models. With many nuances, sure )))
** [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk I think your distinction between "interactive" models is very important!!! See for example the interactionist paradigm of computing by Peter Wegner and Dinah Goldin.

* [[RaviSharma|Ravi Sharma]]: Anatoly would you call L4 Contemplative?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: No. All models about activity (enactive cognition, activity is everywhere).

* [[RaviSharma|Ravi Sharma]]: Generation Differentiation is described, how do you integrate to get all facts together that means knowledge?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: I simply deny all guesses that give problems in inference. All survived guesses are integrated (i.e., not give errors in inference).

* [[RaviSharma|Ravi Sharma]]: Last Q for Anatoly, how do you introduce value systems in these cognitive architectures other than through social media etc.

* [[DouglasMiles|Douglas Miles]]: The most interesting and confusing thing about LLMs is we have no idea how to teach them any new skills... other than: We fire a hose of data and text at them and just pray

* [[AmitSheth|Amit Sheth]]: About passing Turing Test: Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as You May Think -- Introducing AI Detectability Index, https://lnkd.in/gr6cizEZ
** [[ArunMajumdar|Arun Majumdar]]: The distinction between "human" and machine is being attacked on stylometric principles but this is very hard as computer GPT outputs are mixed with human inputs ...

* [[RaviSharma|Ravi Sharma]]: What efforts are likely to succeed in making ChatGPT more accurate?
** [[ArunMajumdar|Arun Majumdar]]: @Ravi Sharma so our approach is to use Conceptual Graphs as our formal knowledge graph approach to creating a "surrogate" model to drive the LLM/GPTs.

* [[ArunMajumdar|Arun Majumdar]]: We (John and I) are not focused on the detection problem because that is something that is not related to our primary focus in knowledge graphs (Conceptual Graphs) as a formalism for symbolic AI with Generative AI.
** [[ArunMajumdar|Arun Majumdar]]: I can share screenshots that compares what we do to what others do. Basically, zero-hallucinations. We can look at a simple hallucination problem later after John's talk if needed. Ken has my screenshots.

* [[RaviSharma|Ravi Sharma]]: Is math not derived from metaphysics?
** [[ArunMajumdar|Arun Majumdar]]: Math is one of the great abstractions of human kind that is a formal but extremely open field for creativity. The key is that creativity in math is often understated. Closed World Models (CWMs) which include LLMs are not capable of creating outside something that they have been input (in their ML training, even with human reinforcement feedback learning).
** [[ArunMajumdar|Arun Majumdar]]: Tuning a GPT or LLM is still a closed world model.
** [[ArunMajumdar|Arun Majumdar]]: Semiotic models differ - the semiotics of Saussure are different to Pierce. One is dyadic, and the other is triadic. So the semiotics matter. These distinctions are not involved in GPT/LLM constructs. However, this important distinction may be approachable in the future.
** [[ArunMajumdar|Arun Majumdar]]: The patterns in mathematics are not probabilistic. They are driven by rational principles. So these are different things.
** [[ArunMajumdar|Arun Majumdar]]: You can use probability, like you can use a million monkeys typing with probabilistic bias to get some pattern candidates but there is no deep insight or principle driving these outputs.

* [[RaviSharma|Ravi Sharma]]: LLM and cognitive scientists and cog-memory well explained.

* [[JanetSinger|Janet Singer]]: Question for Anatoly re 4E cognition. Shift to 4E seems to be the key for appreciating and taking advantage of human-machine teaming opportunities post LLM revolution. But do you see 4E as applicable to software AIs separated from humans? Or just to robots (arguably embodied and embedded if in limited sense)? Or is it that we are the 4E extensions of the software AIs?

* [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk Behavior is not truly just a random process. There is "intention" in living beings - such as the need to survive and thrive. Stochastics are a way to study behavior, to mimic some behaviors but the complexity of goal-driven and intentful behavior is actually much more complex than probabilities.

* [[ArunMajumdar|Arun Majumdar]]: Permion uses tensor mathematics but in the logical formalisms so that predications and symbolic logic has a mapping to/from from the tensorial structures. There are several emerging developments including, for example, the exploration of alternatives algebras such as Clifford or Geometric algebras beyond the conventional real valued Gibbs algebras used nowadays.

* [[DavidEddy|David Eddy]]: 1.5M LoC is NOT a big system

* [[DouglasMiles|Douglas Miles]]: The earth may not be round enough to have a completely circular cord
** [[DouglasMiles|Douglas Miles]]: It takes a person that had experience beyond words to answer that
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a thought experiment discussed by John Sowa - about a circular cord around the earth's equator, and adding 1 yard to it. It raises the cord to approx 6" above the ground.

* [[MikeBennett|Mike Bennett]]: For those of us not from the US, what is the capital of Alaska? It feels like this is a vital plot point.
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a ChatGPT response in John's/Arun's presentation - that Juneau is the capital of Alaska, but there have been attempts to move the capital to another city.

* [[RaviSharma|Ravi Sharma]]: Arun Say more about scaffolding. Is it only E and R and detailed parsed later, but using what rules?
** [[ArunMajumdar|Arun Majumdar]]: Scaffolding is based on purely mathematical methods based on the Zipf distribution laws of terminologies and the H-Point of the Zipf distribution and Eigen computation methods to identify the modules of graph structures from the text. Scaffolding provides a language
agnostic foundation.

* [[DaleFitch|Dale Fitch]]: Is your "logic" an ontology?
** [[ArunMajumdar|Arun Majumdar]]: Yes we induce and also human update the "ontology" but the logic is using the ontology

* [[DavidEddy|David Eddy]]: Thank you Ken, John, Arun & Anatoly.

== Resources ==
* [https://bit.ly/3QMfqZj Video Recording]
* [https://youtu.be/G4S-Zrc5qUk YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 08

2023-11-09T17:14:29Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Broader thoughts]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::8 Nov 2023 17:00 GMT]]
|-
| 9:00am PST/12:00pm EST
|-
| 5:00pm GMT/6:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', strategist and blogger at [http://ailev.lievjournal.ru Laboratory Log]
** ''Knowledge graphs and large language models in cognitive architectures''
** This talk discusses styles of definitions for knowledge graphs (KG) combined with large language models (LLMs). The KG architectures and systems are reviewed, taken from Ontolog Forum's 2020 Communique. A framework is proposed for a cognitive architecture using both LLMa and KGs for the evolution of knowledge during 4E (embodied, extended, embedded, enacted) cognition. In this framework, ontologies are understood as answers to the question "What is in the world?" and can be found in representations that vary across a spectrum of formality/rigor. An example is given of the use of ontology engineering training in management, where upper-level ontologies are given to students in the form of informal course texts (with the goal of obtaining a fine-tuned LLM within the "neural networks" of students' brains) coupled with lower-level ontologies that are more formal (such as data schemas for databases and knowledge graphs).
** Anatoly Levenchuk has worked as a strategy consultant for more than 30 years. He helps with vision and strategy definition to many government agencies and large companies. Now he is science head of Aisystant that serves as a school in engineering and management. His first machine learning project was in 1977, first ontology engineering project was in 1980. He is author of several textbooks on systems thinking, methodology, systems engineering, systems management, natural and artificial intelligence, education as "person engineering". His blog "Laboratory Log" http://ailev.lievjournal.ru in Russian has more than 3,000 subscribers.
** [https://bit.ly/3sljmXt Slides]
* '''[[ArunMajumdar|Arun Majumdar]]''' and '''[[JohnSowa|John Sowa]]''', [https://permion.ai/ Permion AI]
** ''Trustworthy Computation: Diagrammatic Reasoning With and About LLMs''
** Large Language Models (LLMs) were designed for machine translation (MT). Although LLM methods cannot do any reasoning by themselves, they can often find and apply reasoning patterns that they find in the vast resources of the WWW. For common problems, they frequently find a correct solution. For more complex problems, they may construct a solution that is partially correct for some applications, but disastrously wrong or even hallucinogenic for others. Systems developed by Permion use LLMs for what they do best. But they combine them with precise and trusted methods of diagrammatic reasoning based on conceptual graphs (CGs). They take advantage of the full range of technology developed by 60+ years of AI, computer science, and computational linguistics. For any application, Permion methods derive an ontology tailored to the policies, rules, and specifications of the project or business. All programs and results they produce are guaranteed to be consistent with that ontology.
** [https://bit.ly/464bRlF John's Slides]
** [https://bit.ly/475rpH1 Arun's Slides]
* [https://bit.ly/3QMfqZj Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 8 November 2023'''
* Start Time: 9:00am PST / 12:00pm EST / 6:00pm CET / 5:00pm GMT / 1700 UTC
** Note that Daylight Saving Time has ended in Europe, Canada and the US.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=8&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: When I think about models, it is not necessarily graph or even video but something like a kind of vision mind-based understanding and yes I express it often in language or express math using the language also.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Yes, any form. But if it (model) has patterns then you can consider it as text -- semiotics tell us "all is texts of patterns as letters". Pattern languages are about behavior as a text (chain of patterns). Mathematics is about patterns too.

* [[MichaelDeBellis|Michael DeBellis]]: I don't understand what the speaker means by quantum/digital memory.
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: It is about exact copy of information, information is about difference. 1 bit is a result of measurement. To evolution can proceed, you need genes as exactly copied information about results of previous evolution steps. If you have not digital/quantum/discrete form for exact copying, you cannot accumulate knowledge, error in analog form will be prohibitive for evolution.

* [[RaviSharma|Ravi Sharma]]: Second bullet can you expand how cognition and quantum memory are understood?
** [[MikeBennett|Mike Bennett]]: That is the 2nd bullet of Slide 3 (for later reference in the Q&A)
** [[AndreaWesterinen|Andrea Westerinen]]: This is discussed at the end of the talk, in the Q&A

* [[RaviSharma|Ravi Sharma]]: Generative and interactive models, how will these be integrated at different levels?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: Generative and discriminative (not interactive) models. You can ask about any models, but get different types of answers: possible worlds descriptions from generative models and classification labels from discriminative models. With many nuances, sure )))
** [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk I think your distinction between "interactive" models is very important!!! See for example the interactionist paradigm of computing by Peter Wegner and Dinah Goldin.

* [[RaviSharma|Ravi Sharma]]: Anatoly would you call L4 Contemplative?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: No. All models about activity (enactive cognition, activity is everywhere).

* [[RaviSharma|Ravi Sharma]]: Generation Differentiation is described, how do you integrate to get all facts together that means knowledge?
** [[AnatolyLevenchuk|Anatoly Levenchuk]]: I simply deny all guesses that give problems in inference. All survived guesses are integrated (i.e., not give errors in inference).

* [[RaviSharma|Ravi Sharma]]: Last Q for Anatoly, how do you introduce value systems in these cognitive architectures other than through social media etc.

* [[DouglasMiles|Douglas Miles]]: The most interesting and confusing thing about LLMs is we have no idea how to teach them any new skills... other than: We fire a hose of data and text at them and just pray

* [[AmitSheth|Amit Sheth]]: About passing Turing Test: Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as You May Think -- Introducing AI Detectability Index, https://lnkd.in/gr6cizEZ
** [[ArunMajumdar|Arun Majumdar]]: The distinction between "human" and machine is being attacked on stylometric principles but this is very hard as computer GPT outputs are mixed with human inputs ...

* [[RaviSharma|Ravi Sharma]]: What efforts are likely to succeed in making ChatGPT more accurate?
** [[ArunMajumdar|Arun Majumdar]]: @Ravi Sharma so our approach is to use Conceptual Graphs as our formal knowledge graph approach to creating a "surrogate" model to drive the LLM/GPTs.

* [[ArunMajumdar|Arun Majumdar]]: We (John and I) are not focused on the detection problem because that is something that is not related to our primary focus in knowledge graphs (Conceptual Graphs) as a formalism for symbolic AI with Generative AI.
** [[ArunMajumdar|Arun Majumdar]]: I can share screenshots that compares what we do to what others do. Basically, zero-hallucinations. We can look at a simple hallucination problem later after John's talk if needed. Ken has my screenshots.

* [[RaviSharma|Ravi Sharma]]: Is math not derived from metaphysics?
** [[ArunMajumdar|Arun Majumdar]]: Math is one of the great abstractions of human kind that is a formal but extremely open field for creativity. The key is that creativity in math is often understated. Closed World Models (CWMs) which include LLMs are not capable of creating outside something that they have been input (in their ML training, even with human reinforcement feedback learning).
** [[ArunMajumdar|Arun Majumdar]]: Tuning a GPT or LLM is still a closed world model.
** [[ArunMajumdar|Arun Majumdar]]: Semiotic models differ - the semiotics of Saussure are different to Pierce. One is dyadic, and the other is triadic. So the semiotics matter. These distinctions are not involved in GPT/LLM constructs. However, this important distinction may be approachable in the future.
** [[ArunMajumdar|Arun Majumdar]]: The patterns in mathematics are not probabilistic. They are driven by rational principles. So these are different things.
** [[ArunMajumdar|Arun Majumdar]]: You can use probability, like you can use a million monkeys typing with probabilistic bias to get some pattern candidates but there is no deep insight or principle driving these outputs.

* [[RaviSharma|Ravi Sharma]]: LLM and cognitive scientists and cog-memory well explained.

* [[JanetSinger|Janet Singer]]: Question for Anatoly re 4E cognition. Shift to 4E seems to be the key for appreciating and taking advantage of human-machine teaming opportunities post LLM revolution. But do you see 4E as applicable to software AIs separated from humans? Or just to robots (arguably embodied and embedded if in limited sense)? Or is it that we are the 4E extensions of the software AIs?

* [[ArunMajumdar|Arun Majumdar]]: @Anatoly Levenchuk Behavior is not truly just a random process. There is "intention" in living beings - such as the need to survive and thrive. Stochastics are a way to study behavior, to mimic some behaviors but the complexity of goal-driven and intentful behavior is actually much more complex than probabilities.

* [[ArunMajumdar|Arun Majumdar]]: Permion uses tensor mathematics but in the logical formalisms so that predications and symbolic logic has a mapping to/from from the tensorial structures. There are several emerging developments including, for example, the exploration of alternatives algebras such as Clifford or Geometric algebras beyond the conventional real valued Gibbs algebras used nowadays.

* [[DavidEddy|David Eddy]]: 1.5M LoC is NOT a big system

* [[DouglasMiles|Douglas Miles]]: The earth may not be round enough to have a completely circular cord
** [[DouglasMiles|Douglas Miles]]: It takes a person that had experience beyond words to answer that
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a thought experiment discussed by John Sowa - about a circular cord around the earth's equator, and adding 1 yard to it. It raises the cord to approx 6" above the ground.

* [[MikeBennett|Mike Bennett]]: For those of us not from the US, what is the capital of Alaska? It feels like this is a vital plot point.
** [[AndreaWesterinen|Andrea Westerinen]]: This is a reference to a ChatGPT response in John's/Arun's presentation - that Juneau is the capital of Alaska, but there have been attempts to move the capital to another city.

* [[RaviSharma|Ravi Sharma]]: Arun Say more about scaffolding. Is it only E and R and detailed parsed later, but using what rules?
** [[ArunMajumdar|Arun Majumdar]]: Scaffolding is based on purely mathematical methods based on the Zipf distribution laws of terminologies and the H-Point of the Zipf distribution and Eigen computation methods to identify the modules of graph structures from the text. Scaffolding provides a language
agnostic foundation.

* [[DaleFitch|Dale Fitch]]: Is your "logic" an ontology?
** [[ArunMajumdar|Arun Majumdar]]: Yes we induce and also human update the "ontology" but the logic is using the ontology

* [[DavidEddy|David Eddy]]: Thank you Ken, John, Arun & Anatoly.

== Resources ==
* [https://bit.ly/3QMfqZj Video Recording]
* [https://youtu.be/G4S-Zrc5qUk YouTube Video]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_08]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

AndreaWesterinen

2023-11-09T16:38:39Z

AndreaWesterinen:

= Andrea Westerinen =

Consultant and Researcher
'''OntoInsights, LLC'''

email: arwesterinen [at] gmail.com

email: andrea [at] ontoinsights.com

see: https://organizingknowledge.blogspot.com

see: https://hearing-all-voices.blogspot.com

see: http://www.ontoinsights.com

Andrea Westerinen is an independent software engineer and systems architect, and CTO of OntoInsights, LLC (http://ontoinsights.com). She specializes in ontology development and knowledge engineering, and has extensive software development experience. Ms. Westerinen has strong interests in semantic and linguistics technologies, and has worked in the computer industry since 1979, at places like Raytheon/BBN, Two Six Labs, SAIC, CA Technologies, Microsoft, Cisco, Intel and IBM. Her responsibilities have included researcher, strategist, program manager, personnel manager, software developer, ontologist and enthusiast, as needed. Ms. Westerinen has led and participated in many network and systems/storage management standards organizations, and held the positions of Technical Director for the Storage Networking Industry Association (SNIA) and Vice President of Technology for the Distributed Management Task Force (DMTF). She has a B.S. in Physics and Mathematics from Marquette University, and an M.S. in Computer Science from Nova Southeastern University. Ms. Westerinen currently holds 6 patents.

[[Category:Person]]

AndreaWesterinen

2023-11-09T16:38:10Z

AndreaWesterinen:

= Andrea Westerinen =

Consultant and Researcher
'''OntoInsights, LLC'''

email: arwesterinen [at] gmail.com
email: andrea [at] ontoinsights.com

see: https://organizingknowledge.blogspot.com

see: https://hearing-all-voices.blogspot.com

see: http://www.ontoinsights.com

Andrea Westerinen is an independent software engineer and systems architect, and CTO of OntoInsights, LLC (http://ontoinsights.com). She specializes in ontology development and knowledge engineering, and has extensive software development experience. Ms. Westerinen has strong interests in semantic and linguistics technologies, and has worked in the computer industry since 1979, at places like Raytheon/BBN, Two Six Labs, SAIC, CA Technologies, Microsoft, Cisco, Intel and IBM. Her responsibilities have included researcher, strategist, program manager, personnel manager, software developer, ontologist and enthusiast, as needed. Ms. Westerinen has led and participated in many network and systems/storage management standards organizations, and held the positions of Technical Director for the Storage Networking Industry Association (SNIA) and Vice President of Technology for the Distributed Management Task Force (DMTF). She has a B.S. in Physics and Mathematics from Marquette University, and an M.S. in Computer Science from Nova Southeastern University. Ms. Westerinen currently holds 6 patents.

[[Category:Person]]

ConferenceCall 2023 11 08

2023-11-06T19:18:07Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 11 08

2023-11-06T01:03:22Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 11 08

2023-11-06T01:01:32Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 11 01

2023-11-04T18:12:55Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Demos of information extraction via hybrid systems]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::1 Nov
2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CET
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* [[AndreaWesterinen|Andrea Westerinen]], Creator of [https://github.com/ontoinsights/deep_narrative_analysis DNA, Deep Narrative Analysis]
** '''Title:''' Populating Knowledge Graphs: The Confluence of Ontology and Large Language Models
** '''Abstract:''' Ontology-based Knowledge Graphs (KGs) stand at the forefront of semantic data representation, providing structured views of the data in complex domains. Traditionally, populating these KGs from unstructured text involved convoluted natural language analyses and custom code, but the environment has changed with the use of Large Language Models (LLMs). This talk explores one use case - the population of a KG from news articles. The evolution of the DNA application from employing spaCy APIs to OpenAI is described, and the current (open-source) implementation discussed. Implementation issues such as sourcing the data, LLM prompts, mapping the LLM responses onto the ontology, and populating the knowledge graph are overviewed.
** [https://bit.ly/49tTuJY Slides]
* [[PrasadYalamanchi|Prasad Yalamanchi]], [https://leadsemantics.com/ Lead Semantics] CTO
** '''Title:''' Harvest Knowledge From Language - Harness the power of Large Language Models and Semantic Technology
** '''Abstract:''' Language (both text and voice) holds much of the accessible knowledge to humans. It is also the best store of the collective human knowledge. Historically, accessing this knowledge, was manual and up until recent times has progressed to varying degrees of semi-automated methods! But, with the advent of Language Models and particularly Large Language Models in the last couple of years, a fully automated access to knowledge carried in language is now becoming a reality! TextDistil, the software product from Lead Semantics applies LLMs and Ontologies to extract computable knowledge in the form of RDF triples from Text.
** [https://bit.ly/3tY3niI Slides]
* [https://bit.ly/45Z1fUY Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 1 November 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 5:00pm CET / 4:00pm GMT / 1600 UTC
** Note that Daylight Saving Time has ended in Europe but not in the US or Canada.
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=11&day=1&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[PrasadYalamanchi|Prasad Yalamanchi]]
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* [[SusanneVejdemo|Susanne Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* [[JanetSinger|Janet Singer]]
* [[DouglasMiles|Douglas Miles]]
* [[JohnSowa|John Sowa]]
* [[BartGajderowicz|Bart Gajderowicz]]
* [[GaryBergCross|Gary Berg-Cross]]
* Zefi Kavvadia
* ariusz (Telicent)
* Dan (Telicent)
* [[MarkUnderwood|Mark Underwood]]
* Sundos Al Subhi
* Jeff
* James Logan

Complete list of participants was not captured

== Discussion ==
* [[RaviSharma|Ravi Sharma]]: What is the normal accuracy and does the accuracy of triples being unique from language?
** [[RaviSharma|Ravi Sharma]]: Namely would it just depend on the language only or on domain concept would affect the KG?

* [[MichaelDeBellis|Michael DeBellis]]: Is the ontology COMPLETELY created from the Corpus, or do you start from a foundation ontology and extend it based on the Corpus?

* [[RaviSharma|Ravi Sharma]]: You implied preprocessing and semantic understanding by humans before the KG is generated? how much effort is it?

* [[RaviSharma|Ravi Sharma]]: Is there a possibility to reduce the duration by compromising the accuracy somewhat?

* [[SusanneVejdemo|Susanne Vejdemo]]: I understand we have a KG constructed from the unstructured docs. And then there’s translation of your query to triples? I am a bit uncertain where the LLM comes into this?
** [[JanetSinger|Janet Singer]]: My question as well — how exactly does the LLM come in?
** [[AndreaWesterinen|Andrea Westerinen]]: I will address this in my presentation, but can’t answer for Prasad.
** [[PrasadYalamanchi|Prasad Yalamanchi]]: LLM is coming in multiple places in the TextDistil pipeline. Once at the final summary string of the result items. It comes in building the KG, as well
** [[BartGajderowicz|Bart Gajderowicz]]: Prasad, what does the LLM do in building the KG?

* [[RaviSharma|Ravi Sharma]]: Can it show the visuals during progress such as the KG?

* [[DouglasMiles|Douglas Miles]]: Chatgpt-3.5 in some instances can work well enough to be used ovber ChatGPT-4 ?
** [[AndreaWesterinen|Andrea Westerinen]]: It MAY, but I have found profound differences. Linguistic analysis is much better in 4
** [[DouglasMiles|Douglas Miles]]: ChatGPT-3.5 can be so much faster with its return results.. I've considered running both to see when 3.5 was sufficient.. admnittely mostly it isn't.. but "convert this to owl" often is "acceptable"
** [[AndreaWesterinen|Andrea Westerinen]]: @Douglas Miles Not sure that I agree about acceptability.
** [[DouglasMiles|Douglas Miles]]: ok true.. I mean 3.5 cant even begin to convert to CLIF of CycL .. whereas it at least tried with RDF/OWL

* [[RaviSharma|Ravi Sharma]]: If there are no human interventions, how much is the KG affected?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi, that's my question as well. IMO it is usually better to have a person in the loop because creating a well designed ontology completely from a Corpus seems like the resulting ontology may not be well designed. That's why I asked the question about starting from a basic ontology and then extending that ontology, rather than creating the entire ontology from scratch

* [[DouglasMiles|Douglas Miles]]: Prasad presentation was very awesome!

* [[RaviSharma|Ravi Sharma]]: Prasad - If you do two such exercises, is the result the same/repeatable?
** [[PrasadYalamanchi|Prasad Yalamanchi]]: There is a language interpreatation to map the query string to the Ontology.
** [[PrasadYalamanchi|Prasad Yalamanchi]]: SO, if two queries (exercises as you mentioned) result in the same interpretation, then the final answers will be the same

* [[DouglasMiles|Douglas Miles]]: Something that impresses me and is unique about Andrea's work (even year or two ago.. ) ... She actually supports full modality representations in RDF-ish languages.. Stuff that normally I would only dare to use CLIF to represent!

* [[RaviSharma|Ravi Sharma]]: Are there similarities to the rhetoric possibilities of metaphor, context, explanation, etc to improve your results?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure what is being asked. I am exposing the use of rhetorical devices to help readers understand how the text might be affecting their interpretations of it.

* [[RaviSharma|Ravi Sharma]]: Andrea does ML or AI enter this exercise? And results you showed, if so where?
* [[RaviSharma|Ravi Sharma]]: I mean what ML and learning sets were used in OpenAI
** [[AndreaWesterinen|Andrea Westerinen]]: OpenAI's complete technology stack is not disclosed but their website says "We build our generative models using a technology called deep learning, which leverages large amounts of data to train an AI system to perform a task."

* [[DouglasMiles|Douglas Miles]]: To help answer how LLMs can be useful in translation: https://chat.openai.com/share/039d72c3-8432-48d1-98b8-63e15614bbef

* [[MarkUnderwood|Mark Underwood]]: Excellent presentations and important work for the ontology community

* Sundos Al Subhi: Thank you all!! Great information.

* [[JanetSinger|Janet Singer]]: Excellent presentations — Looking forward to seeing these ideas integrated in the future session(s)

* [[DouglasMiles|Douglas Miles]]: (there is no question that the KR Andrea is doing is rock solid!) Here is my question though: Are any of the RDF reasoners good enough to do the reasoning/query that Andrea expects?
** [[AndreaWesterinen|Andrea Westerinen]]: @Douglas Miles Yes, I use Stardog. Also allows use of Voicebox which encode NL queries in SPARQL!

* [[DouglasMiles|Douglas Miles]]: Thank you that was great!

* Dan (Telicent): Thank you for your presentations 🙂

* Zefi Kavvadia : thank you!

* Mariusz (Telicent) : Thank you.

== Resources ==
* [https://bit.ly/45Z1fUY Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_11_01]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_11_01]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 11 01

2023-11-04T17:44:05Z

AndreaWesterinen: /* Participants */

ConferenceCall 2023 11 01

2023-11-04T17:42:25Z

AndreaWesterinen: /* Participants */

ConferenceCall 2023 10 25

2023-11-04T17:36:32Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 2]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::25 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Evren Sirin''', Stardog CTO and lead for their new [https://www.stardog.com/categories/voicebox/ Voicebox] offering
** '''Title:''' Stardog Voicebox: LLM-Powered Question Answering with Knowledge Graphs
** '''Abstract:''' Large Language Models (LLMs) and Generative AI technologies have caused a shift in all areas of information technology but especially for question answering use cases. Leveraging LLMs for question answering can help fully democratize enterprise analytics and data access. However, using LLMs with enterprise data bring significant challenges around security, privacy, accuracy, and explainabilty. In this talk we will present Stardog [https://www.stardog.com/categories/voicebox/ Voicebox] which leverages an open-source foundational LLM to build, manage, and query knowledge graphs using ordinary language. The answers to user questions directly come from the knowledge graph providing complete traceability and access control. Stardog Voicebox combines statistical reasoning in the form of LLMs with logical reasoning in knowledge graphs providing a powerful hybrid reasoning system with a natural language interface.
** [https://bit.ly/3MeGrSy Slides]
* '''Yuan He''', Key contributor to [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a package for ontology engineering with deep learning
** '''Title:''' DeepOnto: A Python Package for Ontology Engineering with Deep Learning and Language Models
** '''Abstract:''' Integrating deep learning techniques, particularly language models (LMs), with knowledge representations like ontologies has raised widespread attention, urging the need for a platform that supports both paradigms. However, deep learning frameworks like PyTorch and Tensorflow are predominantly developed for Python programming, while widely-used ontology APIs, such as the OWL API and Jena, are primarily Java-based. To facilitate seamless integration of these frameworks and APIs, we present [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a Python package designed for ontology engineering with deep learning. The package encompasses a core ontology processing module founded on the widely-recognized and reliable OWL API, encapsulating its fundamental features in a more “Pythonic” manner and extending its capabilities to incorporate other essential components including reasoning, verbalization, normalization, projection, taxonomy, and more. Building on this module, DeepOnto offers a suite of tools, resources, and algorithms that support various ontology engineering tasks, such as ontology alignment and completion, by harnessing deep learning methods, primarily pre-trained LMs.
** [https://bit.ly/46VaEOu Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 25 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=25&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* Evren Sirin
* Yuan He
* Jiaoyan Chen
* Hang Dong
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* Riley Moher
* [[ToddSchneider|Todd Schneider]]
* [[PeteRivett|Pete Rivett]]
* [[DouglasMiles|Douglas Miles]]
* [[JohnSowa|John Sowa]]
* [[BartGajderowicz|Bart Gajderowicz]]
* E S
* [[GaryBergCross|Gary Berg-Cross]]
* silke
* James Logan

Complete list of participants was not captured

== Discussion ==
* (Question before chat recording) How is reasoning supported in Stardog?
** [[AndreaWesterinen|Andrea Westerinen]]: Stardog includes a reasoner. (One of the precursors to Stardog was the Pellet reasoner, which could be used in Protege.)

* [[RaviSharma|Ravi Sharma]]: How does stardog distinguish the contention between datastores APIs Vs App API's differences? Namely distinguishing any discrepencies?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure that I understand your question. The datastore APIs are REST based.

* [[AndreaWesterinen|Andrea Westerinen]]: Making SPARQL more user friendly (NLP -> SPARQL) is valuable.
** [[MichaelDeBellis|Michael DeBellis]]: Andrea, can you say more about NLP -> SPARQL? Is this a new spec, a book,??? Have a link?
** [[MichaelDeBellis|Michael DeBellis]]: Sorry, I misread your comment. I thought you said "is available" not "is valuable" and thought it was some new paper or spec. Anyway, yes I agree this is critical to making Semantic Web get traction in the real world and shouldn't be too difficult. Taking natural language and generating SPARQL from it.
** [[AndreaWesterinen|Andrea Westerinen]]: This is what Evren will talk about ... Voicebox.

* [[AndreaWesterinen|Andrea Westerinen]]: Knowledge comes from KGs only, but NLP is for user-assistance.

* [[AndreaWesterinen|Andrea Westerinen]]: Evren is discussing one example of Stardog knowledge kits - related to beers, ingredients and customers.
** [[AndreaWesterinen|Andrea Westerinen]]: Plus there are inference rules that can also be used in a query.
** Riley Moher: So are we generating a new relation whose sort constraints are determined based on semantic similarity of existing relations?
** [[AndreaWesterinen|Andrea Westerinen]]: You can do a PATH analysis, but this is a discussion of an inference rule that I believe is pre-defined.

* [[ToddSchneider|Todd Schneider]]: When asking for a customer's supplier, what are they supplying?
** [[AndreaWesterinen|Andrea Westerinen]]: Any ingredient used in a product purchased by a customer
** [[AndreaWesterinen|Andrea Westerinen]]: Kind of weird, but I think that the point is using inference to help with NLP translation

* [[MichaelDeBellis|Michael DeBellis]]: I would like to get more detail on how you integrated vector DB with triplestore

* [[PeteRivett|Pete Rivett]]: What's the language used to express the rules?
** [[ToddSchneider|Todd Schneider]]: The ‘rules’ look like SWRL rules (i.e. Horn clauses).
** [[AndreaWesterinen|Andrea Westerinen]]: That is what Stardog uses. Yes.

* [[DouglasMiles|Douglas Miles]]: What opensrc LLMs did you find adequately trained?
** [[AndreaWesterinen|Andrea Westerinen]]: Evren is reporting use of MPT-30B trained by MosaicML/Databricks
** [[MichaelDeBellis|Michael DeBellis]]: For open source LLMs the best place to look IMO is: https://huggingface.co/
** [[DouglasMiles|Douglas Miles]]: So you tried several openSrc Models and only one was barely adequate?!
** Evren Sirin: This was based on an analysis of smaller models (7B). And, this is an ongoing process.
** [[DouglasMiles|Douglas Miles]]: I am doing some NL<->CLIF and i so far only found ChatGPT-4 adequate (CahtGPT-3.5 was absolutely a waste of time).. This is good news if MPT-30B is worth a try for this.
** Riley Moher: Very interested in NL <-> CLIF , what is the nature of the work?
** Evren Sirin: LLM stage is changing rapidly. There are some newcomers like Mistral that is promising. We’ve gotten comparable results with Llama 2 as well.
** [[DouglasMiles|Douglas Miles]]: I wont be surpised if Llama X or whichever will be as good as ChatGPT-4 w/in a short time from now
** [[DouglasMiles|Douglas Miles]]: Just a small amount of seeing how good it is showing someone they probably want to use CLIF over the overly popular ARM for KR
** [[DouglasMiles|Douglas Miles]]: I had lots of good experience with CLIF.. It is at least expressive enough for English... whereas ARM seems not to be
** [[JohnSowa|John Sowa]]: CLIF is much easier to map to and from English than OWL.
** Riley Moher: More expressive for sure
** [[JohnSowa|John Sowa]]: CLIF includes full FOL plus a version of HOL.

* [[ToddSchneider|Todd Schneider]]: The results of computing an inference may or may not be materialized.
** Evren Sirin: Yes, in theory inferences can be materialized but in Stardog we only support query-time inferencing by rewriting the user query and executing with a Datalog engine.

* [[PeteRivett|Pete Rivett]]: Have you considered using an enterprise-specific KG to train a specific LLM?
** Evren Sirin: Yes, this is definitely on our roadmap. We are starting with a general-purpose LLM that can be used with any KG but more domain-specific fine-tuning will certainly improve results.

* [[RaviSharma|Ravi Sharma]]: Example of Vector Embeddings?
** Yuan He: SentenceBERT / FAISS, for example.
** Evren Sirin: We use MiniLM which is a small language model (or more correctly a sentence transformer) for creating vectors from text.

* [[BartGajderowicz|Bart Gajderowicz]]: Can you expand on the traceability aspect?
** Evren Sirin: I tried to showcase this at the end a little. We can see the query used, the data sources and data elements that contributed to the answer, etc. Traceability and explainability is still very low-level (requires RDF and SPARQL knowledge) but that’s something we plan to tackle later.

* [[AndreaWesterinen|Andrea Westerinen]]: What happens if the concept is not known to the KG?
** Riley Moher: What if the natural language query cannot be expressed with SPARQL?
** Evren Sirin: At this stage we simple say question cannot be answered. Our goal is to use the conversational aspects to clarify the question and/or clarify to the user that graph does not contain relevant information.

* E S: @Evren Sirin Supply Chain usage demonstration is very useful and I think applicable in business. What are the required specs for hardware for Stardog? Would you please share the estimated costs for monthly fixed costs to maintain the whole system, ie regardless of customers usage? Thank you.
** Evren Sirin: There are lots of different considerations that would go into hardware specs. We typically suggest people to start with out hosted option that has a Free tier:https://www.stardog.com/stardog-cloud/ For on-prem deployment, there is capacity planning discussion here https://docs.stardog.com/operating-stardog/server-administration/capacity-planning

* Jiaoyan Chen: We are happy to answer questions on DeepOnto in the ChatBox.

* [[RaviSharma|Ravi Sharma]]: Yuan Great start, on ontology engineering

* [[AndreaWesterinen|Andrea Westerinen]]: Where was the FoodProduct rule defined?

* [[ToddSchneider|Todd Schneider]]: Explainability is a huge benefit over LLMs.

* [[RaviSharma|Ravi Sharma]]: I see verification and logical merging, embedding etc.

* [[ToddSchneider|Todd Schneider]]: Yuan, please explain ‘subsumption restructuring’.1
* Jiaoyan Chen: I do not fully capture in which slide this phrase happens. If it is in ontology-to-graph, it refers to our work that extracting a class hierarchy from the original ontology. There is special case: A \equiv B \conjunciton C, we will have A \subclassof B and A \subclassof C; this is to avoid placing A under owl:Thing, if we just consider the declared subsumptions of named classes for building the hierarchy. This is also the strategy of Protege for pressing the class hierarchy.

* [[RaviSharma|Ravi Sharma]]: What happens if semantics from same entities have differences?
* Hang Dong: One idea is to do automated concept discovery and insertion (from texts for example). We are still exploring to implement these in DeepOnto. One recent work https://arxiv.org/abs/2306.14704 and https://arxiv.org/abs/2302.07189

* [[GaryBergCross|Gary Berg-Cross]]: @Yuan Can you say a bit more about how alignment between ontologies is done?
* Jiaoyan Chen : Briefly, it fine-tunes a BERT-based binary classifier with synonyms from the ontologies to be aligned, uses the classifier to predict candidate equivalent class pairs with class labels, combines the prediction scores with lexical matching scores, and finally uses logical reasoning for consistency checking and repair (using a repairing algorithm our group developed before).
** [[RaviSharma|Ravi Sharma]]: In that case how is alignment taken care of?

* [[MichaelDeBellis|Michael DeBellis]]: Both of these presentations were excellent! Very useful info.

* [[RaviSharma|Ravi Sharma]]: Can you tell more about OAEI?
** Jiaoyan Chen: https://www.cs.ox.ac.uk/isg/projects/ConCur/oaei/2023/index.html
** Hang Dong: It is an onto matching benchmarking activity running for many years.
** Jiaoyan Chen: We placed a new Bio-ML track in OAEI which has been made for over a decade.
** Jiaoyan Chen: Our new Bio-ML track was place in 2022, and is continuing in 2023. This track is especially developed for ML-based OM systems.

* [[JohnSowa|John Sowa]]: If you use LLMs to do translation from English to CLIF, the mapping is simpler and more successful.

* silke: Can you please give a short explanation of logic repair and how it is implemented? Thanks!
** Jiaoyan Chen: Yes. We get mappings and their scores. Briefly, the repair algorithm merges the mappings and the ontologies to infer whether they are consistent. If not, it tries to remove some mappings with lowest scores, and see whether the remaining mappings + the ontologies are consistent. If yes, it stops. This procedure is iterative. The reasoning is approximated using Propositional logics.
** Jiaoyan Chen: More details are here: https://ceur-ws.org/Vol-1014/paper_63.pdf
** silke: Thank you so much, very helpful!

* [[JohnSowa|John Sowa]]: For any mapping from NL to any other notation, you need an "echo".
** [[JohnSowa|John Sowa]]: Whenever you type anything in English, the system should produce an echo in English to show exactly how your input was interpreted.
** [[DouglasMiles|Douglas Miles]]: I have asked it to translate the CLIF back to English and then told it to tell me if the original English matches it response and to modify the English->CLIF .. tis 2nd round produces much better results
** [[JohnSowa|John Sowa]]: If the echo is not what you wanted, you can revise your question.

* James LOGAN: Can DeepOnto create axioms that seem to always hold true in some domain from a text corpus?
** Jiaoyan Chen: Not yet. We now are trying to extract new concepts from text and insert them into the ontology (there are some ongoing works: https://arxiv.org/abs/2306.14704. We haven’t consider axioms, but only concepts. It’s a good idea for the future extension.

* [[ToddSchneider|Todd Schneider]]: Please explain ‘subsumption restructuring’.
** Yuan He: We introduced subsumption axioms between parents and children concepts of a concept target for removal.
** James LOGAN: It seems this would require jumping to conclusions or having a way to close the world

* [[MikeBennett|Mike Bennett]]: Yuan is the ontology alignment etc. dependent on any specific TLO? Or can different TLOs be used as the basis for this?
** Jiaoyan Chen: I don’t know what’s TLO, but I think not …
** [[MikeBennett|Mike Bennett]]: Top Level Ontology
** Yuan He: Just depend on the input ontologies is sufficient.
** Jiaoyan Chen: No, it does not
** [[ToddSchneider|Todd Schneider]]: ‘Top Level Ontology’ equivalent to ‘Foundational Ontology’

* E S: Is there any accuracy problem in building KG with other languages than English?
** Jiaoyan Chen: DeepOnto currently is tested only for English ontologies
** [[AndreaWesterinen|Andrea Westerinen]]: You need the appropriate training for LLMs. So, you have translation and translation errors.

== Resources ==
* [https://bit.ly/3MeGrSy Evrin Sirin Slides]
* [https://bit.ly/46VaEOu Yuan He Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 25

2023-11-04T17:36:15Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 2]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::25 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Evren Sirin''', Stardog CTO and lead for their new [https://www.stardog.com/categories/voicebox/ Voicebox] offering
** '''Title:''' Stardog Voicebox: LLM-Powered Question Answering with Knowledge Graphs
** '''Abstract:''' Large Language Models (LLMs) and Generative AI technologies have caused a shift in all areas of information technology but especially for question answering use cases. Leveraging LLMs for question answering can help fully democratize enterprise analytics and data access. However, using LLMs with enterprise data bring significant challenges around security, privacy, accuracy, and explainabilty. In this talk we will present Stardog [https://www.stardog.com/categories/voicebox/ Voicebox] which leverages an open-source foundational LLM to build, manage, and query knowledge graphs using ordinary language. The answers to user questions directly come from the knowledge graph providing complete traceability and access control. Stardog Voicebox combines statistical reasoning in the form of LLMs with logical reasoning in knowledge graphs providing a powerful hybrid reasoning system with a natural language interface.
** [https://bit.ly/3MeGrSy Slides]
* '''Yuan He''', Key contributor to [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a package for ontology engineering with deep learning
** '''Title:''' DeepOnto: A Python Package for Ontology Engineering with Deep Learning and Language Models
** '''Abstract:''' Integrating deep learning techniques, particularly language models (LMs), with knowledge representations like ontologies has raised widespread attention, urging the need for a platform that supports both paradigms. However, deep learning frameworks like PyTorch and Tensorflow are predominantly developed for Python programming, while widely-used ontology APIs, such as the OWL API and Jena, are primarily Java-based. To facilitate seamless integration of these frameworks and APIs, we present [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a Python package designed for ontology engineering with deep learning. The package encompasses a core ontology processing module founded on the widely-recognized and reliable OWL API, encapsulating its fundamental features in a more “Pythonic” manner and extending its capabilities to incorporate other essential components including reasoning, verbalization, normalization, projection, taxonomy, and more. Building on this module, DeepOnto offers a suite of tools, resources, and algorithms that support various ontology engineering tasks, such as ontology alignment and completion, by harnessing deep learning methods, primarily pre-trained LMs.
** [https://bit.ly/46VaEOu Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 25 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=25&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
Complete list of participants was not captured

* Evren Sirin
* Yuan He
* Jiaoyan Chen
* Hang Dong
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* Riley Moher
* [[ToddSchneider|Todd Schneider]]
* [[PeteRivett|Pete Rivett]]
* [[DouglasMiles|Douglas Miles]]
* [[JohnSowa|John Sowa]]
* [[BartGajderowicz|Bart Gajderowicz]]
* E S
* [[GaryBergCross|Gary Berg-Cross]]
* silke
* James Logan

== Discussion ==
* (Question before chat recording) How is reasoning supported in Stardog?
** [[AndreaWesterinen|Andrea Westerinen]]: Stardog includes a reasoner. (One of the precursors to Stardog was the Pellet reasoner, which could be used in Protege.)

* [[RaviSharma|Ravi Sharma]]: How does stardog distinguish the contention between datastores APIs Vs App API's differences? Namely distinguishing any discrepencies?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure that I understand your question. The datastore APIs are REST based.

* [[AndreaWesterinen|Andrea Westerinen]]: Making SPARQL more user friendly (NLP -> SPARQL) is valuable.
** [[MichaelDeBellis|Michael DeBellis]]: Andrea, can you say more about NLP -> SPARQL? Is this a new spec, a book,??? Have a link?
** [[MichaelDeBellis|Michael DeBellis]]: Sorry, I misread your comment. I thought you said "is available" not "is valuable" and thought it was some new paper or spec. Anyway, yes I agree this is critical to making Semantic Web get traction in the real world and shouldn't be too difficult. Taking natural language and generating SPARQL from it.
** [[AndreaWesterinen|Andrea Westerinen]]: This is what Evren will talk about ... Voicebox.

* [[AndreaWesterinen|Andrea Westerinen]]: Knowledge comes from KGs only, but NLP is for user-assistance.

* [[AndreaWesterinen|Andrea Westerinen]]: Evren is discussing one example of Stardog knowledge kits - related to beers, ingredients and customers.
** [[AndreaWesterinen|Andrea Westerinen]]: Plus there are inference rules that can also be used in a query.
** Riley Moher: So are we generating a new relation whose sort constraints are determined based on semantic similarity of existing relations?
** [[AndreaWesterinen|Andrea Westerinen]]: You can do a PATH analysis, but this is a discussion of an inference rule that I believe is pre-defined.

* [[ToddSchneider|Todd Schneider]]: When asking for a customer's supplier, what are they supplying?
** [[AndreaWesterinen|Andrea Westerinen]]: Any ingredient used in a product purchased by a customer
** [[AndreaWesterinen|Andrea Westerinen]]: Kind of weird, but I think that the point is using inference to help with NLP translation

* [[MichaelDeBellis|Michael DeBellis]]: I would like to get more detail on how you integrated vector DB with triplestore

* [[PeteRivett|Pete Rivett]]: What's the language used to express the rules?
** [[ToddSchneider|Todd Schneider]]: The ‘rules’ look like SWRL rules (i.e. Horn clauses).
** [[AndreaWesterinen|Andrea Westerinen]]: That is what Stardog uses. Yes.

* [[DouglasMiles|Douglas Miles]]: What opensrc LLMs did you find adequately trained?
** [[AndreaWesterinen|Andrea Westerinen]]: Evren is reporting use of MPT-30B trained by MosaicML/Databricks
** [[MichaelDeBellis|Michael DeBellis]]: For open source LLMs the best place to look IMO is: https://huggingface.co/
** [[DouglasMiles|Douglas Miles]]: So you tried several openSrc Models and only one was barely adequate?!
** Evren Sirin: This was based on an analysis of smaller models (7B). And, this is an ongoing process.
** [[DouglasMiles|Douglas Miles]]: I am doing some NL<->CLIF and i so far only found ChatGPT-4 adequate (CahtGPT-3.5 was absolutely a waste of time).. This is good news if MPT-30B is worth a try for this.
** Riley Moher: Very interested in NL <-> CLIF , what is the nature of the work?
** Evren Sirin: LLM stage is changing rapidly. There are some newcomers like Mistral that is promising. We’ve gotten comparable results with Llama 2 as well.
** [[DouglasMiles|Douglas Miles]]: I wont be surpised if Llama X or whichever will be as good as ChatGPT-4 w/in a short time from now
** [[DouglasMiles|Douglas Miles]]: Just a small amount of seeing how good it is showing someone they probably want to use CLIF over the overly popular ARM for KR
** [[DouglasMiles|Douglas Miles]]: I had lots of good experience with CLIF.. It is at least expressive enough for English... whereas ARM seems not to be
** [[JohnSowa|John Sowa]]: CLIF is much easier to map to and from English than OWL.
** Riley Moher: More expressive for sure
** [[JohnSowa|John Sowa]]: CLIF includes full FOL plus a version of HOL.

* [[ToddSchneider|Todd Schneider]]: The results of computing an inference may or may not be materialized.
** Evren Sirin: Yes, in theory inferences can be materialized but in Stardog we only support query-time inferencing by rewriting the user query and executing with a Datalog engine.

* [[PeteRivett|Pete Rivett]]: Have you considered using an enterprise-specific KG to train a specific LLM?
** Evren Sirin: Yes, this is definitely on our roadmap. We are starting with a general-purpose LLM that can be used with any KG but more domain-specific fine-tuning will certainly improve results.

* [[RaviSharma|Ravi Sharma]]: Example of Vector Embeddings?
** Yuan He: SentenceBERT / FAISS, for example.
** Evren Sirin: We use MiniLM which is a small language model (or more correctly a sentence transformer) for creating vectors from text.

* [[BartGajderowicz|Bart Gajderowicz]]: Can you expand on the traceability aspect?
** Evren Sirin: I tried to showcase this at the end a little. We can see the query used, the data sources and data elements that contributed to the answer, etc. Traceability and explainability is still very low-level (requires RDF and SPARQL knowledge) but that’s something we plan to tackle later.

* [[AndreaWesterinen|Andrea Westerinen]]: What happens if the concept is not known to the KG?
** Riley Moher: What if the natural language query cannot be expressed with SPARQL?
** Evren Sirin: At this stage we simple say question cannot be answered. Our goal is to use the conversational aspects to clarify the question and/or clarify to the user that graph does not contain relevant information.

* E S: @Evren Sirin Supply Chain usage demonstration is very useful and I think applicable in business. What are the required specs for hardware for Stardog? Would you please share the estimated costs for monthly fixed costs to maintain the whole system, ie regardless of customers usage? Thank you.
** Evren Sirin: There are lots of different considerations that would go into hardware specs. We typically suggest people to start with out hosted option that has a Free tier:https://www.stardog.com/stardog-cloud/ For on-prem deployment, there is capacity planning discussion here https://docs.stardog.com/operating-stardog/server-administration/capacity-planning

* Jiaoyan Chen: We are happy to answer questions on DeepOnto in the ChatBox.

* [[RaviSharma|Ravi Sharma]]: Yuan Great start, on ontology engineering

* [[AndreaWesterinen|Andrea Westerinen]]: Where was the FoodProduct rule defined?

* [[ToddSchneider|Todd Schneider]]: Explainability is a huge benefit over LLMs.

* [[RaviSharma|Ravi Sharma]]: I see verification and logical merging, embedding etc.

* [[ToddSchneider|Todd Schneider]]: Yuan, please explain ‘subsumption restructuring’.1
* Jiaoyan Chen: I do not fully capture in which slide this phrase happens. If it is in ontology-to-graph, it refers to our work that extracting a class hierarchy from the original ontology. There is special case: A \equiv B \conjunciton C, we will have A \subclassof B and A \subclassof C; this is to avoid placing A under owl:Thing, if we just consider the declared subsumptions of named classes for building the hierarchy. This is also the strategy of Protege for pressing the class hierarchy.

* [[RaviSharma|Ravi Sharma]]: What happens if semantics from same entities have differences?
* Hang Dong: One idea is to do automated concept discovery and insertion (from texts for example). We are still exploring to implement these in DeepOnto. One recent work https://arxiv.org/abs/2306.14704 and https://arxiv.org/abs/2302.07189

* [[GaryBergCross|Gary Berg-Cross]]: @Yuan Can you say a bit more about how alignment between ontologies is done?
* Jiaoyan Chen : Briefly, it fine-tunes a BERT-based binary classifier with synonyms from the ontologies to be aligned, uses the classifier to predict candidate equivalent class pairs with class labels, combines the prediction scores with lexical matching scores, and finally uses logical reasoning for consistency checking and repair (using a repairing algorithm our group developed before).
** [[RaviSharma|Ravi Sharma]]: In that case how is alignment taken care of?

* [[MichaelDeBellis|Michael DeBellis]]: Both of these presentations were excellent! Very useful info.

* [[RaviSharma|Ravi Sharma]]: Can you tell more about OAEI?
** Jiaoyan Chen: https://www.cs.ox.ac.uk/isg/projects/ConCur/oaei/2023/index.html
** Hang Dong: It is an onto matching benchmarking activity running for many years.
** Jiaoyan Chen: We placed a new Bio-ML track in OAEI which has been made for over a decade.
** Jiaoyan Chen: Our new Bio-ML track was place in 2022, and is continuing in 2023. This track is especially developed for ML-based OM systems.

* [[JohnSowa|John Sowa]]: If you use LLMs to do translation from English to CLIF, the mapping is simpler and more successful.

* silke: Can you please give a short explanation of logic repair and how it is implemented? Thanks!
** Jiaoyan Chen: Yes. We get mappings and their scores. Briefly, the repair algorithm merges the mappings and the ontologies to infer whether they are consistent. If not, it tries to remove some mappings with lowest scores, and see whether the remaining mappings + the ontologies are consistent. If yes, it stops. This procedure is iterative. The reasoning is approximated using Propositional logics.
** Jiaoyan Chen: More details are here: https://ceur-ws.org/Vol-1014/paper_63.pdf
** silke: Thank you so much, very helpful!

* [[JohnSowa|John Sowa]]: For any mapping from NL to any other notation, you need an "echo".
** [[JohnSowa|John Sowa]]: Whenever you type anything in English, the system should produce an echo in English to show exactly how your input was interpreted.
** [[DouglasMiles|Douglas Miles]]: I have asked it to translate the CLIF back to English and then told it to tell me if the original English matches it response and to modify the English->CLIF .. tis 2nd round produces much better results
** [[JohnSowa|John Sowa]]: If the echo is not what you wanted, you can revise your question.

* James LOGAN: Can DeepOnto create axioms that seem to always hold true in some domain from a text corpus?
** Jiaoyan Chen: Not yet. We now are trying to extract new concepts from text and insert them into the ontology (there are some ongoing works: https://arxiv.org/abs/2306.14704. We haven’t consider axioms, but only concepts. It’s a good idea for the future extension.

* [[ToddSchneider|Todd Schneider]]: Please explain ‘subsumption restructuring’.
** Yuan He: We introduced subsumption axioms between parents and children concepts of a concept target for removal.
** James LOGAN: It seems this would require jumping to conclusions or having a way to close the world

* [[MikeBennett|Mike Bennett]]: Yuan is the ontology alignment etc. dependent on any specific TLO? Or can different TLOs be used as the basis for this?
** Jiaoyan Chen: I don’t know what’s TLO, but I think not …
** [[MikeBennett|Mike Bennett]]: Top Level Ontology
** Yuan He: Just depend on the input ontologies is sufficient.
** Jiaoyan Chen: No, it does not
** [[ToddSchneider|Todd Schneider]]: ‘Top Level Ontology’ equivalent to ‘Foundational Ontology’

* E S: Is there any accuracy problem in building KG with other languages than English?
** Jiaoyan Chen: DeepOnto currently is tested only for English ontologies
** [[AndreaWesterinen|Andrea Westerinen]]: You need the appropriate training for LLMs. So, you have translation and translation errors.

== Resources ==
* [https://bit.ly/3MeGrSy Evrin Sirin Slides]
* [https://bit.ly/46VaEOu Yuan He Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 25

2023-11-04T17:35:24Z

AndreaWesterinen:

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 2]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::25 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Evren Sirin''', Stardog CTO and lead for their new [https://www.stardog.com/categories/voicebox/ Voicebox] offering
** '''Title:''' Stardog Voicebox: LLM-Powered Question Answering with Knowledge Graphs
** '''Abstract:''' Large Language Models (LLMs) and Generative AI technologies have caused a shift in all areas of information technology but especially for question answering use cases. Leveraging LLMs for question answering can help fully democratize enterprise analytics and data access. However, using LLMs with enterprise data bring significant challenges around security, privacy, accuracy, and explainabilty. In this talk we will present Stardog [https://www.stardog.com/categories/voicebox/ Voicebox] which leverages an open-source foundational LLM to build, manage, and query knowledge graphs using ordinary language. The answers to user questions directly come from the knowledge graph providing complete traceability and access control. Stardog Voicebox combines statistical reasoning in the form of LLMs with logical reasoning in knowledge graphs providing a powerful hybrid reasoning system with a natural language interface.
** [https://bit.ly/3MeGrSy Slides]
* '''Yuan He''', Key contributor to [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a package for ontology engineering with deep learning
** '''Title:''' DeepOnto: A Python Package for Ontology Engineering with Deep Learning and Language Models
** '''Abstract:''' Integrating deep learning techniques, particularly language models (LMs), with knowledge representations like ontologies has raised widespread attention, urging the need for a platform that supports both paradigms. However, deep learning frameworks like PyTorch and Tensorflow are predominantly developed for Python programming, while widely-used ontology APIs, such as the OWL API and Jena, are primarily Java-based. To facilitate seamless integration of these frameworks and APIs, we present [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a Python package designed for ontology engineering with deep learning. The package encompasses a core ontology processing module founded on the widely-recognized and reliable OWL API, encapsulating its fundamental features in a more “Pythonic” manner and extending its capabilities to incorporate other essential components including reasoning, verbalization, normalization, projection, taxonomy, and more. Building on this module, DeepOnto offers a suite of tools, resources, and algorithms that support various ontology engineering tasks, such as ontology alignment and completion, by harnessing deep learning methods, primarily pre-trained LMs.
** [https://bit.ly/46VaEOu Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 25 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=25&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* Evren Sirin
* Yuan He
* Jiaoyan Chen
* Hang Dong
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[MichaelDeBellis|Michael DeBellis]]
* Riley Moher
* [[ToddSchneider|Todd Schneider]]
* [[PeteRivett|Pete Rivett]]
* [[DouglasMiles|Douglas Miles]]
* [[JohnSowa|John Sowa]]
* [[BartGajderowicz|Bart Gajderowicz]]
* E S
* [[GaryBergCross|Gary Berg-Cross]]
* silke
* James Logan

== Discussion ==
* (Question before chat recording) How is reasoning supported in Stardog?
** [[AndreaWesterinen|Andrea Westerinen]]: Stardog includes a reasoner. (One of the precursors to Stardog was the Pellet reasoner, which could be used in Protege.)

* [[RaviSharma|Ravi Sharma]]: How does stardog distinguish the contention between datastores APIs Vs App API's differences? Namely distinguishing any discrepencies?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure that I understand your question. The datastore APIs are REST based.

* [[AndreaWesterinen|Andrea Westerinen]]: Making SPARQL more user friendly (NLP -> SPARQL) is valuable.
** [[MichaelDeBellis|Michael DeBellis]]: Andrea, can you say more about NLP -> SPARQL? Is this a new spec, a book,??? Have a link?
** [[MichaelDeBellis|Michael DeBellis]]: Sorry, I misread your comment. I thought you said "is available" not "is valuable" and thought it was some new paper or spec. Anyway, yes I agree this is critical to making Semantic Web get traction in the real world and shouldn't be too difficult. Taking natural language and generating SPARQL from it.
** [[AndreaWesterinen|Andrea Westerinen]]: This is what Evren will talk about ... Voicebox.

* [[AndreaWesterinen|Andrea Westerinen]]: Knowledge comes from KGs only, but NLP is for user-assistance.

* [[AndreaWesterinen|Andrea Westerinen]]: Evren is discussing one example of Stardog knowledge kits - related to beers, ingredients and customers.
** [[AndreaWesterinen|Andrea Westerinen]]: Plus there are inference rules that can also be used in a query.
** Riley Moher: So are we generating a new relation whose sort constraints are determined based on semantic similarity of existing relations?
** [[AndreaWesterinen|Andrea Westerinen]]: You can do a PATH analysis, but this is a discussion of an inference rule that I believe is pre-defined.

* [[ToddSchneider|Todd Schneider]]: When asking for a customer's supplier, what are they supplying?
** [[AndreaWesterinen|Andrea Westerinen]]: Any ingredient used in a product purchased by a customer
** [[AndreaWesterinen|Andrea Westerinen]]: Kind of weird, but I think that the point is using inference to help with NLP translation

* [[MichaelDeBellis|Michael DeBellis]]: I would like to get more detail on how you integrated vector DB with triplestore

* [[PeteRivett|Pete Rivett]]: What's the language used to express the rules?
** [[ToddSchneider|Todd Schneider]]: The ‘rules’ look like SWRL rules (i.e. Horn clauses).
** [[AndreaWesterinen|Andrea Westerinen]]: That is what Stardog uses. Yes.

* [[DouglasMiles|Douglas Miles]]: What opensrc LLMs did you find adequately trained?
** [[AndreaWesterinen|Andrea Westerinen]]: Evren is reporting use of MPT-30B trained by MosaicML/Databricks
** [[MichaelDeBellis|Michael DeBellis]]: For open source LLMs the best place to look IMO is: https://huggingface.co/
** [[DouglasMiles|Douglas Miles]]: So you tried several openSrc Models and only one was barely adequate?!
** Evren Sirin: This was based on an analysis of smaller models (7B). And, this is an ongoing process.
** [[DouglasMiles|Douglas Miles]]: I am doing some NL<->CLIF and i so far only found ChatGPT-4 adequate (CahtGPT-3.5 was absolutely a waste of time).. This is good news if MPT-30B is worth a try for this.
** Riley Moher: Very interested in NL <-> CLIF , what is the nature of the work?
** Evren Sirin: LLM stage is changing rapidly. There are some newcomers like Mistral that is promising. We’ve gotten comparable results with Llama 2 as well.
** [[DouglasMiles|Douglas Miles]]: I wont be surpised if Llama X or whichever will be as good as ChatGPT-4 w/in a short time from now
** [[DouglasMiles|Douglas Miles]]: Just a small amount of seeing how good it is showing someone they probably want to use CLIF over the overly popular ARM for KR
** [[DouglasMiles|Douglas Miles]]: I had lots of good experience with CLIF.. It is at least expressive enough for English... whereas ARM seems not to be
** [[JohnSowa|John Sowa]]: CLIF is much easier to map to and from English than OWL.
** Riley Moher: More expressive for sure
** [[JohnSowa|John Sowa]]: CLIF includes full FOL plus a version of HOL.

* [[ToddSchneider|Todd Schneider]]: The results of computing an inference may or may not be materialized.
** Evren Sirin: Yes, in theory inferences can be materialized but in Stardog we only support query-time inferencing by rewriting the user query and executing with a Datalog engine.

* [[PeteRivett|Pete Rivett]]: Have you considered using an enterprise-specific KG to train a specific LLM?
** Evren Sirin: Yes, this is definitely on our roadmap. We are starting with a general-purpose LLM that can be used with any KG but more domain-specific fine-tuning will certainly improve results.

* [[RaviSharma|Ravi Sharma]]: Example of Vector Embeddings?
** Yuan He: SentenceBERT / FAISS, for example.
** Evren Sirin: We use MiniLM which is a small language model (or more correctly a sentence transformer) for creating vectors from text.

* [[BartGajderowicz|Bart Gajderowicz]]: Can you expand on the traceability aspect?
** Evren Sirin: I tried to showcase this at the end a little. We can see the query used, the data sources and data elements that contributed to the answer, etc. Traceability and explainability is still very low-level (requires RDF and SPARQL knowledge) but that’s something we plan to tackle later.

* [[AndreaWesterinen|Andrea Westerinen]]: What happens if the concept is not known to the KG?
** Riley Moher: What if the natural language query cannot be expressed with SPARQL?
** Evren Sirin: At this stage we simple say question cannot be answered. Our goal is to use the conversational aspects to clarify the question and/or clarify to the user that graph does not contain relevant information.

* E S: @Evren Sirin Supply Chain usage demonstration is very useful and I think applicable in business. What are the required specs for hardware for Stardog? Would you please share the estimated costs for monthly fixed costs to maintain the whole system, ie regardless of customers usage? Thank you.
** Evren Sirin: There are lots of different considerations that would go into hardware specs. We typically suggest people to start with out hosted option that has a Free tier:https://www.stardog.com/stardog-cloud/ For on-prem deployment, there is capacity planning discussion here https://docs.stardog.com/operating-stardog/server-administration/capacity-planning

* Jiaoyan Chen: We are happy to answer questions on DeepOnto in the ChatBox.

* [[RaviSharma|Ravi Sharma]]: Yuan Great start, on ontology engineering

* [[AndreaWesterinen|Andrea Westerinen]]: Where was the FoodProduct rule defined?

* [[ToddSchneider|Todd Schneider]]: Explainability is a huge benefit over LLMs.

* [[RaviSharma|Ravi Sharma]]: I see verification and logical merging, embedding etc.

* [[ToddSchneider|Todd Schneider]]: Yuan, please explain ‘subsumption restructuring’.1
* Jiaoyan Chen: I do not fully capture in which slide this phrase happens. If it is in ontology-to-graph, it refers to our work that extracting a class hierarchy from the original ontology. There is special case: A \equiv B \conjunciton C, we will have A \subclassof B and A \subclassof C; this is to avoid placing A under owl:Thing, if we just consider the declared subsumptions of named classes for building the hierarchy. This is also the strategy of Protege for pressing the class hierarchy.

* [[RaviSharma|Ravi Sharma]]: What happens if semantics from same entities have differences?
* Hang Dong: One idea is to do automated concept discovery and insertion (from texts for example). We are still exploring to implement these in DeepOnto. One recent work https://arxiv.org/abs/2306.14704 and https://arxiv.org/abs/2302.07189

* [[GaryBergCross|Gary Berg-Cross]]: @Yuan Can you say a bit more about how alignment between ontologies is done?
* Jiaoyan Chen : Briefly, it fine-tunes a BERT-based binary classifier with synonyms from the ontologies to be aligned, uses the classifier to predict candidate equivalent class pairs with class labels, combines the prediction scores with lexical matching scores, and finally uses logical reasoning for consistency checking and repair (using a repairing algorithm our group developed before).
** [[RaviSharma|Ravi Sharma]]: In that case how is alignment taken care of?

* [[MichaelDeBellis|Michael DeBellis]]: Both of these presentations were excellent! Very useful info.

* [[RaviSharma|Ravi Sharma]]: Can you tell more about OAEI?
** Jiaoyan Chen: https://www.cs.ox.ac.uk/isg/projects/ConCur/oaei/2023/index.html
** Hang Dong: It is an onto matching benchmarking activity running for many years.
** Jiaoyan Chen: We placed a new Bio-ML track in OAEI which has been made for over a decade.
** Jiaoyan Chen: Our new Bio-ML track was place in 2022, and is continuing in 2023. This track is especially developed for ML-based OM systems.

* [[JohnSowa|John Sowa]]: If you use LLMs to do translation from English to CLIF, the mapping is simpler and more successful.

* silke: Can you please give a short explanation of logic repair and how it is implemented? Thanks!
** Jiaoyan Chen: Yes. We get mappings and their scores. Briefly, the repair algorithm merges the mappings and the ontologies to infer whether they are consistent. If not, it tries to remove some mappings with lowest scores, and see whether the remaining mappings + the ontologies are consistent. If yes, it stops. This procedure is iterative. The reasoning is approximated using Propositional logics.
** Jiaoyan Chen: More details are here: https://ceur-ws.org/Vol-1014/paper_63.pdf
** silke: Thank you so much, very helpful!

* [[JohnSowa|John Sowa]]: For any mapping from NL to any other notation, you need an "echo".
** [[JohnSowa|John Sowa]]: Whenever you type anything in English, the system should produce an echo in English to show exactly how your input was interpreted.
** [[DouglasMiles|Douglas Miles]]: I have asked it to translate the CLIF back to English and then told it to tell me if the original English matches it response and to modify the English->CLIF .. tis 2nd round produces much better results
** [[JohnSowa|John Sowa]]: If the echo is not what you wanted, you can revise your question.

* James LOGAN: Can DeepOnto create axioms that seem to always hold true in some domain from a text corpus?
** Jiaoyan Chen: Not yet. We now are trying to extract new concepts from text and insert them into the ontology (there are some ongoing works: https://arxiv.org/abs/2306.14704. We haven’t consider axioms, but only concepts. It’s a good idea for the future extension.

* [[ToddSchneider|Todd Schneider]]: Please explain ‘subsumption restructuring’.
** Yuan He: We introduced subsumption axioms between parents and children concepts of a concept target for removal.
** James LOGAN: It seems this would require jumping to conclusions or having a way to close the world

* [[MikeBennett|Mike Bennett]]: Yuan is the ontology alignment etc. dependent on any specific TLO? Or can different TLOs be used as the basis for this?
** Jiaoyan Chen: I don’t know what’s TLO, but I think not …
** [[MikeBennett|Mike Bennett]]: Top Level Ontology
** Yuan He: Just depend on the input ontologies is sufficient.
** Jiaoyan Chen: No, it does not
** [[ToddSchneider|Todd Schneider]]: ‘Top Level Ontology’ equivalent to ‘Foundational Ontology’

* E S: Is there any accuracy problem in building KG with other languages than English?
** Jiaoyan Chen: DeepOnto currently is tested only for English ontologies
** [[AndreaWesterinen|Andrea Westerinen]]: You need the appropriate training for LLMs. So, you have translation and translation errors.

== Resources ==
* [https://bit.ly/3MeGrSy Evrin Sirin Slides]
* [https://bit.ly/46VaEOu Yuan He Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 25

2023-11-04T17:29:48Z

AndreaWesterinen:

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 2]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::25 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Evren Sirin''', Stardog CTO and lead for their new [https://www.stardog.com/categories/voicebox/ Voicebox] offering
** '''Title:''' Stardog Voicebox: LLM-Powered Question Answering with Knowledge Graphs
** '''Abstract:''' Large Language Models (LLMs) and Generative AI technologies have caused a shift in all areas of information technology but especially for question answering use cases. Leveraging LLMs for question answering can help fully democratize enterprise analytics and data access. However, using LLMs with enterprise data bring significant challenges around security, privacy, accuracy, and explainabilty. In this talk we will present Stardog [https://www.stardog.com/categories/voicebox/ Voicebox] which leverages an open-source foundational LLM to build, manage, and query knowledge graphs using ordinary language. The answers to user questions directly come from the knowledge graph providing complete traceability and access control. Stardog Voicebox combines statistical reasoning in the form of LLMs with logical reasoning in knowledge graphs providing a powerful hybrid reasoning system with a natural language interface.
** [https://bit.ly/3MeGrSy Slides]
* '''Yuan He''', Key contributor to [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a package for ontology engineering with deep learning
** '''Title:''' DeepOnto: A Python Package for Ontology Engineering with Deep Learning and Language Models
** '''Abstract:''' Integrating deep learning techniques, particularly language models (LMs), with knowledge representations like ontologies has raised widespread attention, urging the need for a platform that supports both paradigms. However, deep learning frameworks like PyTorch and Tensorflow are predominantly developed for Python programming, while widely-used ontology APIs, such as the OWL API and Jena, are primarily Java-based. To facilitate seamless integration of these frameworks and APIs, we present [https://krr-oxford.github.io/DeepOnto/ DeepOnto], a Python package designed for ontology engineering with deep learning. The package encompasses a core ontology processing module founded on the widely-recognized and reliable OWL API, encapsulating its fundamental features in a more “Pythonic” manner and extending its capabilities to incorporate other essential components including reasoning, verbalization, normalization, projection, taxonomy, and more. Building on this module, DeepOnto offers a suite of tools, resources, and algorithms that support various ontology engineering tasks, such as ontology alignment and completion, by harnessing deep learning methods, primarily pre-trained LMs.
** [https://bit.ly/46VaEOu Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 25 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=25&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==

== Discussion ==
* (Question before chat recording) How is reasoning supported in Stardog?
** [[AndreaWesterinen|Andrea Westerinen]]: Stardog includes a reasoner. (One of the precursors to Stardog was the Pellet reasoner, which could be used in Protege.)

* [[RaviSharma|Ravi Sharma]]: How does stardog distinguish the contention between datastores APIs Vs App API's differences? Namely distinguishing any discrepencies?
** [[AndreaWesterinen|Andrea Westerinen]]: I am not sure that I understand your question. The datastore APIs are REST based.

* [[AndreaWesterinen|Andrea Westerinen]]: Making SPARQL more user friendly (NLP -> SPARQL) is valuable.
** [[MichaelDeBellis|Michael DeBellis]]: Andrea, can you say more about NLP -> SPARQL? Is this a new spec, a book,??? Have a link?
** [[MichaelDeBellis|Michael DeBellis]]: Sorry, I misread your comment. I thought you said "is available" not "is valuable" and thought it was some new paper or spec. Anyway, yes I agree this is critical to making Semantic Web get traction in the real world and shouldn't be too difficult. Taking natural language and generating SPARQL from it.
** [[AndreaWesterinen|Andrea Westerinen]]: This is what Evren will talk about ... Voicebox.

* [[AndreaWesterinen|Andrea Westerinen]]: Knowledge comes from KGs only, but NLP is for user-assistance.

* [[AndreaWesterinen|Andrea Westerinen]]: Evren is discussing one example of Stardog knowledge kits - related to beers, ingredients and customers.
** [[AndreaWesterinen|Andrea Westerinen]]: Plus there are inference rules that can also be used in a query.
** Riley Moher: So are we generating a new relation whose sort constraints are determined based on semantic similarity of existing relations?
** [[AndreaWesterinen|Andrea Westerinen]]: You can do a PATH analysis, but this is a discussion of an inference rule that I believe is pre-defined.

* [[ToddSchneider|Todd Schneider]]: When asking for a customer's supplier, what are they supplying?
** [[AndreaWesterinen|Andrea Westerinen]]: Any ingredient used in a product purchased by a customer
** [[AndreaWesterinen|Andrea Westerinen]]: Kind of weird, but I think that the point is using inference to help with NLP translation

* [[MichaelDeBellis|Michael DeBellis]]: I would like to get more detail on how you integrated vector DB with triplestore

* [[PeteRivett|Pete Rivett]]: What's the language used to express the rules?
** [[ToddSchneider|Todd Schneider]]: The ‘rules’ look like SWRL rules (i.e. Horn clauses).
** [[AndreaWesterinen|Andrea Westerinen]]: That is what Stardog uses. Yes.

* [[DouglasMiles|Douglas Miles]]: What opensrc LLMs did you find adequately trained?
** [[AndreaWesterinen|Andrea Westerinen]]: Evren is reporting use of MPT-30B trained by MosaicML/Databricks
** [[MichaelDeBellis|Michael DeBellis]]: For open source LLMs the best place to look IMO is: https://huggingface.co/
** [[DouglasMiles|Douglas Miles]]: So you tried several openSrc Models and only one was barely adequate?!
** Evren Sirin: This was based on an analysis of smaller models (7B). And, this is an ongoing process.
** [[DouglasMiles|Douglas Miles]]: I am doing some NL<->CLIF and i so far only found ChatGPT-4 adequate (CahtGPT-3.5 was absolutely a waste of time).. This is good news if MPT-30B is worth a try for this.
** Riley Moher: Very interested in NL <-> CLIF , what is the nature of the work?
** Evren Sirin: LLM stage is changing rapidly. There are some newcomers like Mistral that is promising. We’ve gotten comparable results with Llama 2 as well.
** [[DouglasMiles|Douglas Miles]]: I wont be surpised if Llama X or whichever will be as good as ChatGPT-4 w/in a short time from now
** [[DouglasMiles|Douglas Miles]]: Just a small amount of seeing how good it is showing someone they probably want to use CLIF over the overly popular ARM for KR
** [[DouglasMiles|Douglas Miles]]: I had lots of good experience with CLIF.. It is at least expressive enough for English... whereas ARM seems not to be
** [[JohnSowa|John Sowa]]: CLIF is much easier to map to and from English than OWL.
** Riley Moher: More expressive for sure
** [[JohnSowa|John Sowa]]: CLIF includes full FOL plus a version of HOL.

* [[ToddSchneider|Todd Schneider]]: The results of computing an inference may or may not be materialized.
** Evren Sirin: Yes, in theory inferences can be materialized but in Stardog we only support query-time inferencing by rewriting the user query and executing with a Datalog engine.

* [[PeteRivett|Pete Rivett]]: Have you considered using an enterprise-specific KG to train a specific LLM?
** Evren Sirin: Yes, this is definitely on our roadmap. We are starting with a general-purpose LLM that can be used with any KG but more domain-specific fine-tuning will certainly improve results.

* [[RaviSharma|Ravi Sharma]]: Example of Vector Embeddings?
** Yuan He: SentenceBERT / FAISS, for example.
** Evren Sirin: We use MiniLM which is a small language model (or more correctly a sentence transformer) for creating vectors from text.

* [[BartGajderowicz|Bart Gajderowicz]]: Can you expand on the traceability aspect?
** Evren Sirin: I tried to showcase this at the end a little. We can see the query used, the data sources and data elements that contributed to the answer, etc. Traceability and explainability is still very low-level (requires RDF and SPARQL knowledge) but that’s something we plan to tackle later.

* [[AndreaWesterinen|Andrea Westerinen]]: What happens if the concept is not known to the KG?
** Riley Moher: What if the natural language query cannot be expressed with SPARQL?
** Evren Sirin: At this stage we simple say question cannot be answered. Our goal is to use the conversational aspects to clarify the question and/or clarify to the user that graph does not contain relevant information.

* E S: @Evren Sirin Supply Chain usage demonstration is very useful and I think applicable in business. What are the required specs for hardware for Stardog? Would you please share the estimated costs for monthly fixed costs to maintain the whole system, ie regardless of customers usage? Thank you.
** Evren Sirin: There are lots of different considerations that would go into hardware specs. We typically suggest people to start with out hosted option that has a Free tier:https://www.stardog.com/stardog-cloud/ For on-prem deployment, there is capacity planning discussion here https://docs.stardog.com/operating-stardog/server-administration/capacity-planning

* Jiaoyan Chen: We are happy to answer questions on DeepOnto in the ChatBox.

* [[RaviSharma|Ravi Sharma]]: Yuan Great start, on ontology engineering

* [[AndreaWesterinen|Andrea Westerinen]]: Where was the FoodProduct rule defined?

* [[ToddSchneider|Todd Schneider]]: Explainability is a huge benefit over LLMs.

* [[RaviSharma|Ravi Sharma]]: I see verification and logical merging, embedding etc.

* [[ToddSchneider|Todd Schneider]]: Yuan, please explain ‘subsumption restructuring’.1
* Jiaoyan Chen: I do not fully capture in which slide this phrase happens. If it is in ontology-to-graph, it refers to our work that extracting a class hierarchy from the original ontology. There is special case: A \equiv B \conjunciton C, we will have A \subclassof B and A \subclassof C; this is to avoid placing A under owl:Thing, if we just consider the declared subsumptions of named classes for building the hierarchy. This is also the strategy of Protege for pressing the class hierarchy.

* [[RaviSharma|Ravi Sharma]]: What happens if semantics from same entities have differences?
* Hang Dong: One idea is to do automated concept discovery and insertion (from texts for example). We are still exploring to implement these in DeepOnto. One recent work https://arxiv.org/abs/2306.14704 and https://arxiv.org/abs/2302.07189

* [[GaryBergCross|Gary Berg-Cross]]: @Yuan Can you say a bit more about how alignment between ontologies is done?
* Jiaoyan Chen : Briefly, it fine-tunes a BERT-based binary classifier with synonyms from the ontologies to be aligned, uses the classifier to predict candidate equivalent class pairs with class labels, combines the prediction scores with lexical matching scores, and finally uses logical reasoning for consistency checking and repair (using a repairing algorithm our group developed before).
** [[RaviSharma|Ravi Sharma]]: In that case how is alignment taken care of?

* [[MichaelDeBellis|Michael DeBellis]]: Both of these presentations were excellent! Very useful info.

* [[RaviSharma|Ravi Sharma]]: Can you tell more about OAEI?
** Jiaoyan Chen: https://www.cs.ox.ac.uk/isg/projects/ConCur/oaei/2023/index.html
** Hang Dong: It is an onto matching benchmarking activity running for many years.
** Jiaoyan Chen: We placed a new Bio-ML track in OAEI which has been made for over a decade.
** Jiaoyan Chen: Our new Bio-ML track was place in 2022, and is continuing in 2023. This track is especially developed for ML-based OM systems.

* [[JohnSowa|John Sowa]]: If you use LLMs to do translation from English to CLIF, the mapping is simpler and more successful.

* silke: Can you please give a short explanation of logic repair and how it is implemented? Thanks!
** Jiaoyan Chen: Yes. We get mappings and their scores. Briefly, the repair algorithm merges the mappings and the ontologies to infer whether they are consistent. If not, it tries to remove some mappings with lowest scores, and see whether the remaining mappings + the ontologies are consistent. If yes, it stops. This procedure is iterative. The reasoning is approximated using Propositional logics.
** Jiaoyan Chen: More details are here: https://ceur-ws.org/Vol-1014/paper_63.pdf
** silke: Thank you so much, very helpful!

* [[JohnSowa|John Sowa]]: For any mapping from NL to any other notation, you need an "echo".
** [[JohnSowa|John Sowa]]: Whenever you type anything in English, the system should produce an echo in English to show exactly how your input was interpreted.
** [[DouglasMiles|Douglas Miles]]: I have asked it to translate the CLIF back to English and then told it to tell me if the original English matches it response and to modify the English->CLIF .. tis 2nd round produces much better results
** [[JohnSowa|John Sowa]]: If the echo is not what you wanted, you can revise your question.

* James LOGAN: Can DeepOnto create axioms that seem to always hold true in some domain from a text corpus?
** Jiaoyan Chen: Not yet. We now are trying to extract new concepts from text and insert them into the ontology (there are some ongoing works: https://arxiv.org/abs/2306.14704. We haven’t consider axioms, but only concepts. It’s a good idea for the future extension.

* [[ToddSchneider|Todd Schneider]]: Please explain ‘subsumption restructuring’.
** Yuan He: We introduced subsumption axioms between parents and children concepts of a concept target for removal.
** James LOGAN: It seems this would require jumping to conclusions or having a way to close the world

* [[MikeBennett|Mike Bennett]]: Yuan is the ontology alignment etc. dependent on any specific TLO? Or can different TLOs be used as the basis for this?
** Jiaoyan Chen: I don’t know what’s TLO, but I think not …
** [[MikeBennett|Mike Bennett]]: Top Level Ontology
** Yuan He: Just depend on the input ontologies is sufficient.
** Jiaoyan Chen: No, it does not
** [[ToddSchneider|Todd Schneider]]: ‘Top Level Ontology’ equivalent to ‘Foundational Ontology’

* E S: Is there any accuracy problem in building KG with other languages than English?
** Jiaoyan Chen: DeepOnto currently is tested only for English ontologies
** [[AndreaWesterinen|Andrea Westerinen]]: You need the appropriate training for LLMs. So, you have translation and translation errors.

== Resources ==
* [https://bit.ly/3MeGrSy Evrin Sirin Slides]
* [https://bit.ly/46VaEOu Yuan He Slides]
* [https://bit.ly/3Q3o0Be Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_25]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 18

2023-11-04T04:26:59Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 1]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::18 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Kurt Cagle''', Author of [https://thecaglereport.com/ The Cagle Report]
** '''Title:''' Complementary Thinking: Language Models, Ontologies and Knowledge Graphs
** '''Abstract:''' With the advent of Retrieval Augmented Generators (RAGs), a more or less standardized workflow has become available for integrating large language models such as ChatGPT with knowledge graphs. This in turn has raised the question about the nature of ontologies associated with LLMs and how knowledge graphs can be structured and queried to make integrated data access possible between the two types of systems. In this talk, Editor and AI Explorer Kurt Cagle of The Cagle Report looks at this process and discusses how they affect both knowledge portals and ontology design.
** [https://bit.ly/3S040lR Slides]
* '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/ LinkedIn])
** '''Title:''' How Ontologies Can Unlock the Potential of Large Language Models for Business
** '''Abstract:''' LLMs have remarkable capabilities; they can craft letters, analyze data, orchestrate workflows, generate code, and much more. Companies such as Google, Apple, Amazon, Meta, and Microsoft are all investing heavily in this technology. Everything indicates that LLMs have enormous disruptive potential. However, there is a problem: they can hallucinate, and for any serious business, that is a deal-breaker. This is where ontologies can come in. In combination with Knowledge Graphs, they can place guardrails around the LLMs, thus allowing organizations to harness the capabilities of LLMs within the framework of a safely controlled ontological structure.
** [https://bit.ly/46A3EH2 Slides]

[https://bit.ly/46DJyvo Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 18 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=18&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
80+ participants

* Kurt Cagle
* Tony Seale
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* Ayya Niyyanika Bhikkhuni
* Anthony Alcaraz
* Alan Morrison
* [[BartGajderowicz|Bart Gajderowicz]]
* Chris Day
* Andrew McCaffrey
* [[GaryBergCross|Gary Berg-Cross]]
* [[JohnSowa|John Sowa]]
* [[MarkUnderwood|Mark Underwood (IS Innovation)]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[SusanneVejdemo|Sus Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* Yishan Liu
* James Logan
* [[PennyAnderson|Penny Anderson]]
* Michael Robbins
* [[MichaelDeBellis|Michael DeBellis]]
* Amit Jain
* Cedric Berger
* Andreas Lothe Opdahl
* Harvey King
* Benoit Claise
* Liju Fan
* Larry Swanson
* Justin Lewis

== Discussion ==
* Anthony's OtterPilot: Hi, I'm an AI assistant helping Anthony Alcaraz take notes for this meeting. Follow along the transcript here: https://otter.ai/u/WLIaj2w-OmOEoVVCP5gURhhZHaY?utm_source=va_chat_link_2  You'll also be able to see screenshots of key moments, add highlights, comments, or action items to anything being said, and get an automatic summary after the meeting.
** [[MikeBennett|Mike Bennett]]: Dang! Put on a session on AI and the AIs start showing up!

* [[AndreaWesterinen|Andrea Westerinen]]: I like the quote, “Ontologies are the shapes of information and knowledge”. Also they provide "information for communication".

* [[ToddSchneider|Todd Schneider]]: What is a ‘shape of information’?
** [[AndreaWesterinen|Andrea Westerinen]]: Roles, constraints, relationships for a domain; a local representation

* [[RaviSharma|Ravi Sharma]]: Can LLMs and/or ontologies be used to detect AI artifacts?
** [[AndreaWesterinen|Andrea Westerinen]]: @Ravi Sharma There are some different articles on this, but basically no.
** [[MikeBennett|Mike Bennett]]: Not unless you could write an ontology that defines what it is to be truly human
** Michael Robbins: Or rebuild the web from the bottom up to embed new frameworks for digital identity and content provenance/authenticity (which is what we need to commit to)
** [[AndreaWesterinen|Andrea Westerinen]]: Provenance also aids with detecting bots

* [[MichaelDeBellis|Michael DeBellis]]: Someone asked about agents in relation to Data Mesh (can't find the orig comment) IMO Data Mesh could be implemented in terms of Agents but typically isn't.

* [[RaviSharma|Ravi Sharma]]: Is there one kind or multiple kinds of connectivity in KG as well as in Ontologies?

* [[RaviSharma|Ravi Sharma]]: Container in the sense of domain overlaps especially overlapping vocabs as Venn diagrams?

* Anh: Why is it hard to build LLMs for other languages? Why can't it be replicated easily when translation work (e.g. Facebook, Google Translate, LinkedIn) has already done a somewhat good job?
** [[AndreaWesterinen|Andrea Westerinen]]: It is about the training data.
** [[BartGajderowicz|Bart Gajderowicz]]: [Anh], on longer text, translation looses many of the cultural nuances of a language, and looses context. Also, most training data is in English so most models are trained on English then translated.
** Anh: Thank you, @Andrea Westerinen & @Bart Gajderowicz. Does it meean it'd cost the same to build a new LLMs for a new language?
** [[AndreaWesterinen|Andrea Westerinen]]: I would believe so.
** [[BartGajderowicz|Bart Gajderowicz]]: The volume of English text for training is cheaper, it’s just the web. So I’d imagine finding enough text in your target language would be the biggest cost. Librarians are our friends here 🙂
** [[PennyAnderson|Penny Anderson]]: this context free is difficult when you start thinking of utility of UI
** Anh: Could you please elaborate more on utility of UI? @Penny Anderson
** [[PennyAnderson|Penny Anderson]]: I mean UIs tend to be process driven data-centric is not tightly bound to a particular process that is context free
** [[PennyAnderson|Penny Anderson]]: other than search ?
** Anh: I see. Thanks.
** [[PennyAnderson|Penny Anderson]]: Data providence, Ethical AI ?
** [[AndreaWesterinen|Andrea Westerinen]]: @Penny Anderson Much better declared and reasoned against in KGs.
** [[PennyAnderson|Penny Anderson]]: Zero trust in networks?

* [[RaviSharma|Ravi Sharma]]: Kurt when you talk of box, you are essentially stating in and out of scope items or is there a way of capturing the info outside the box and bringing it in?

* [[AndreaWesterinen|Andrea Westerinen]]: "What is valid for a graph? Shapes and data"

* [[MichaelDeBellis|Michael DeBellis]]: I don't agree that knowledge graphs are a form of data mesh. They are two related but different concepts. Data mesh is essentially an architecture for enterprise data definition and management. Knowledge graphs are a tool that can be used to implement a data mesh. Data mesh IMO brings the philosophy of microservices to data.
** Alan Morrison: Michael, re: microservices to data, should we be thinking of agents as messengers and KGs as the data resource>

* [[RaviSharma|Ravi Sharma]]: What kind of aggregates are these data shapes? Are these only valid for a class of data or can you mix data types in a shape aggregate?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi: in SHACL you define similar things as you do with OWL. E.g., does a property have to have exactly one value, the datatypes of a property, other constraints. The difference is OWL is used for reasoning over large knowledge graphs and uses the Open World Assumption. SHACL is for constraining data so it uses the Closed World Assumption. E.g., you can define ss_number as a property that must have exactly one value in either OWL or SHACL but in OWL you will almost never trigger an error if the restriction isn't satisfied due to OWA. With SHACL you will get error messages due to CWA

* [[ToddSchneider|Todd Schneider]]: LLM reasoning is probabilistic.

* Ayya Niyyanika Bhikkhuni: When weighting a concept is that some sort of credibility score?

* Michael Robbins: LangChains - I’m thinking about what role ConLangs could serve its intermediaries n revolutionizing language modes and NLP. Happy to have a follow-up discussion with anyone who is interested. https://en.wikipedia.org/wiki/Constructed_language

* [[RaviSharma|Ravi Sharma]]: What is the equivalent of Objects in LLM? What are these entities called and can same onto-entity be different in LLM context?
** [[MikeBennett|Mike Bennett]]: Words, I presume?

* [[RaviSharma|Ravi Sharma]]: Kurt thanks for including wonderful valuable background cultural images, these are inspiring. Are you also conveying the there is external (databased) and internal knowledge such as contemplation?

* [[RaviSharma|Ravi Sharma]]: Why are you limiting your examples to RDF why not MOF also?

* [[ToddSchneider|Todd Schneider]]: The important question is not mapping(s). It’s how can well constructed ontologies be used in the ingestion/training of LLMs.

* [[SusanneVejdemo|Sus Vejdemo]]: Question: what are good case studies of KGE and LLM integrations?
** [[AndreaWesterinen|Andrea Westerinen]]: I addressed some of this in the opening session. “Hybrid systems” include both where LLMs help ontologies (actually the Oct 25th session) and where ontologies help LLMs (Oct 4 and Nov 1 sessions).
** [[AndreaWesterinen|Andrea Westerinen]]: Also, Tony is highlighting a GREAT integration.

* [[MichaelDeBellis|Michael DeBellis]]: One question I have is what happens once you load a knowledge graph into an LLM? I know it can be done but once you load say a Turtle file into the LLM then what?
** [[BartGajderowicz|Bart Gajderowicz]]: Generally the LLM builds a small knowledge graph instance inside its memory and you can query it. Ask it to write SPARQL to get some instances, etc. I have not seen it used for large KGs, just small ones.

* Amit Jain: Will the recording be shared with attendees after the summit?
** [[MikeBennett|Mike Bennett]]: Yes, the recording will be uploaded to the session page when it is ready.

* Cedric Berger: Is there any metrics to measure that indeed KG combines with LLM are less hallucinating?
** [[AndreaWesterinen|Andrea Westerinen]]: I will try to provide these in the summary. I have read papers on this as well as blog posts.
** [[BartGajderowicz|Bart Gajderowicz]]: All the work I’ve come across relies on the knowledge graph to provide explicit knowledge. So if you can ground the LLM with a graph, you can verify if the answers the LLM provides are “facts” in the graph

* Kurt Cagle: Once you load in the ontology into a conversation, it will create (to some extent) an LLM conceptual space for that data. Also keep in mind that getting ALL of a knowledge graph via a RAG is usually not feasible.

* [[RaviSharma|Ravi Sharma]]: Tony great diagram for LLM + ontology to improve each other.why is ontology weak in capturing concepts Vs LLM?

* [[MichaelDeBellis|Michael DeBellis]]: One of the most interesting papers I've read is from Lawrence Berkeley Labs on using LLMs to extend an ontology.

* [[RaviSharma|Ravi Sharma]]: How would models such as in physics work with LLM and Ontology loop or cycle that you show, actually ontologies are conceptually richer than KGs alone?
** [[AndreaWesterinen|Andrea Westerinen]]: KGs are the data, ontologies are the concepts … So, it does not seem right to ask about one being richer.
** Michael Robbins: https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its-wolfram-superpowers/

* [[RaviSharma|Ravi Sharma]]: Tony the built-in uncertainty in LLMs gives it extra power to apply to real life probabilistic world?

* [[RaviSharma|Ravi Sharma]]: Tony LLMs and analog and ontology as Quantum? great way. thanks

* [[MikeBennett|Mike Bennett]]: Left brain v right brain is a great way of talking about ontology v LLM. Now you have to create a good corpus collosum.
** [[GaryBergCross|Gary Berg-Cross]]: A better model than left right hemispheres is by layers - old, mid brain (associative) and neo-cortex. They interconnect in many ways and some by the limbic system.
** [[BartGajderowicz|Bart Gajderowicz]]: I like the System 1 / System 2 analogy

* [[AndreaWesterinen|Andrea Westerinen]]: Tony characterized knowledge graphs as discrete and LLMs as continuous.

* Cedric Berger: Aren’t LLMs also kind of discrete as relying on vectors (arrays of numbers) of limited dimensions?
** [[BartGajderowicz|Bart Gajderowicz]]: LLMs are fundamentally probabilistic, not discrete. Some models are trained to provide discrete classifications, but that’s just at the output level. Internally they are probabilistic.
** [[AndreaWesterinen|Andrea Westerinen]]: The discreteness pertains to the encoding. But, does the probabilistic nature of LLMs make it more continuous? I am not sure.
** [[MikeBennett|Mike Bennett]]: The weightings are a number rather than a logical truth value as in KGs
** Kurt Cagle: Even with KGs, you can set up reifications that also set up Bayesians that are again more fuzzy (or at least more stochastic).
** [[MikeBennett|Mike Bennett]]: If you had an ontology with weightings instead of truth values and can train those values, you have a semantic network like a brain/mind.
** Kurt Cagle: It's where I think we're heading. People in the semantic space have known for years that knowledge is fuzzy / fractal, but getting there has always been the rub.
** [[MikeBennett|Mike Bennett]]: I roughed out an idea for this kind of semantic network application back in the 90s.

* Michael Robbins: A great article on vector embeddings: https://kdb.ai/learning-hub/fundamentals/vector-embeddings/ How can we use this for transparency and explainability? Give users confidence intervals (and other potential response options) along with responses?

* [[RaviSharma|Ravi Sharma]]: What are vectors equivalent to in LLM context?

* Harvey King: Are the embeddings stored across 3 layers?
** [[AndreaWesterinen|Andrea Westerinen]]: The embeddings are there, but simplified/reduced.

* Anh: Does KG consider the time stamp of the assertions/objects? Context of my question: could we use it to mark the originality of posts of similar contents to alert plagiarism.
** [[AndreaWesterinen|Andrea Westerinen]]: The KG CAN do this, if it is encoded.

* [[RaviSharma|Ravi Sharma]]: Has anyone done this a roundtrip quality check, learn from LLM and put it in ontologies and the other way around?
** Benoit Claise: In which context/use case?
** [[MichaelDeBellis|Michael DeBellis]]: That paper I posted earlier from Lawrence Berkeley Labs used LLM to extend an ontology but just went in one direction, expanding the ontology not changing the LLM.

* [[RaviSharma|Ravi Sharma]]: Can you do reasoning on same concept in both to differentiate their respective strengths and weakness?

* [[RaviSharma|Ravi Sharma]]: Tony your tree or chain of thought are great ways

* [[RaviSharma|Ravi Sharma]]: Tony and Kurt Can you address feature space Vs training set learning approaches?

* Liju Fan: Why are the relations in the ontology explicit?
** [[AndreaWesterinen|Andrea Westerinen]]: Relations are defined, and they can be inferred, but this is the essence of ontology. Ontologies are “open world” but do need relationships.
** [[MichaelDeBellis|Michael DeBellis]]: Because in the ontology you (usually manually) create the relations. In an LLM the relations are inferred by the ML algorithm and usually can't be manually changed
** Liju Fan: It seems there is a need to be able to rename LLM inferred relations for them to be human-understandable and practically useful.

* Kurt Cagle: Please note the similarity of Tony's slide with biological cells. Hmmm ...

* Cedric Berger: Has anyone asked AI to generate an image of a factual made of network base units?

* Michael Robbins: Agreed, Tony. And we’ve talked about this on LinkedIn. A constellation of domain-specific and ecosystem-based Community Knowledge Graphs and Language Models. #CKGs and #CLMs
** [[MikeBennett|Mike Bennett]]: TLO as mitochondria?
** Michael Robbins: Language is inseparable from culture and context

* Harvey King: Do KG's take a different nature when dealing with math?
** [[AndreaWesterinen|Andrea Westerinen]]: Not different, but with more rules?
** [[MichaelDeBellis|Michael DeBellis]]: Yes. An ontology uses explicit models like linear algebra. LLMs use linear algebra but computes answers based on examples, doesn't have a theoretical model of math (or other domains) as an ontology does.

* [[SusanneVejdemo|Sus Vejdemo]]: Thanks for a great talk! I love the conceptualization of embeddings as ontologies.

* [[MikeBennett|Mike Bennett]]: If you reify every edge you can give each one an analog value

* [[MarkUnderwood|Mark Underwood]]: For those interested in practical cybersec use cases (lots of chatty network data, some NLP), lot of narrow domain-specific emergent ontologies; e.g., Lambda / microservice mesh etc. mark.underwood@syf.com

* [[GaryBergCross|Gary Berg-Cross]]: I think the working memory graph is a useful early concept bring the LLMs and ontologies together but it is not so easy to capture what is the context for knowledge in this representation. I would guess this is a sub-set of the fluid knowledge of what human cognition employs. Much remains unconscious. But with research AI systems may make more of this explicit.

* [[RaviSharma|Ravi Sharma]]: My query is what is the relationship among reification, provenance and context history?

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Tony, “LLMs for compute” and “As much data into graph, then translating the graph paths to NL and adding to the LLM”

* [[GaryBergCross|Gary Berg-Cross]]: Reality may be atomistically discrete but at such a nano-level that continuous models make better predictions than discrete models that are orders of magnitude too gross rather than fine grained.

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Kurt: “Community Language Models - decentralized, federated, ad hoc network of information”

* [[GaryBergCross|Gary Berg-Cross]]: Models need to be more than federated. Because we center on semantic accuracy and relevance they need to be semantically harmonized.

== Resources ==
* [https://bit.ly/3S040lR Kurt Cagle Slides]
* [https://bit.ly/46A3EH2 Tony Seale Slides]
* [https://bit.ly/46DJyvo Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 18

2023-11-04T01:51:34Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 1]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::18 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Kurt Cagle''', Author of [https://thecaglereport.com/ The Cagle Report]
** '''Title:''' Complementary Thinking: Language Models, Ontologies and Knowledge Graphs
** '''Abstract:''' With the advent of Retrieval Augmented Generators (RAGs), a more or less standardized workflow has become available for integrating large language models such as ChatGPT with knowledge graphs. This in turn has raised the question about the nature of ontologies associated with LLMs and how knowledge graphs can be structured and queried to make integrated data access possible between the two types of systems. In this talk, Editor and AI Explorer Kurt Cagle of The Cagle Report looks at this process and discusses how they affect both knowledge portals and ontology design.
** [https://bit.ly/3S040lR Slides]
* '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/ LinkedIn])
** '''Title:''' How Ontologies Can Unlock the Potential of Large Language Models for Business
** '''Abstract:''' LLMs have remarkable capabilities; they can craft letters, analyze data, orchestrate workflows, generate code, and much more. Companies such as Google, Apple, Amazon, Meta, and Microsoft are all investing heavily in this technology. Everything indicates that LLMs have enormous disruptive potential. However, there is a problem: they can hallucinate, and for any serious business, that is a deal-breaker. This is where ontologies can come in. In combination with Knowledge Graphs, they can place guardrails around the LLMs, thus allowing organizations to harness the capabilities of LLMs within the framework of a safely controlled ontological structure.
** [https://bit.ly/46A3EH2 Slides]

[https://bit.ly/46DJyvo Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 18 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=18&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
80+ participants

* Kurt Cagle
* Tony Seale
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* Ayya Niyyanika Bhikkhuni
* Anthony Alcaraz
* Alan Morrison
* [[BartGajderowicz|Bart Gajderowicz]]
* Chris Day
* Andrew McCaffrey
* [[GaryBergCross|Gary Berg-Cross]]
* [[JohnSowa|John Sowa]]
* [[MarkUnderwood|Mark Underwood (IS Innovation)]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[SusanneVejdemo|Sus Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* Yishan Liu
* James Logan
* [[PennyAnderson|Penny Anderson]]
* Michael Robbins
* [[MichaelDeBellis|Michael DeBellis]]
* Amit Jain
* Cedric Berger
* Andreas Lothe Opdahl
* Harvey King
* Benoit Claise
* Liju Fan
* Larry Swanson
* Justin Lewis

== Discussion ==
* Anthony's OtterPilot: Hi, I'm an AI assistant helping Anthony Alcaraz take notes for this meeting. Follow along the transcript here: https://otter.ai/u/WLIaj2w-OmOEoVVCP5gURhhZHaY?utm_source=va_chat_link_2  You'll also be able to see screenshots of key moments, add highlights, comments, or action items to anything being said, and get an automatic summary after the meeting.
** [[MikeBennett|Mike Bennett]]: Dang! Put on a session on AI and the AIs start showing up!

* [[AndreaWesterinen|Andrea Westerinen]]: I like the quote, “Ontologies are the shapes of information and knowledge”. Also provide "information for communication".

* [[ToddSchneider|Todd Schneider]]: What is a ‘shape of information’?
** [[AndreaWesterinen|Andrea Westerinen]]: Roles, constraints, relationships for a domain; a local representation

* [[RaviSharma|Ravi Sharma]]: Can LLMs and/or ontologies be used to detect AI artifacts?
** [[AndreaWesterinen|Andrea Westerinen]]: @Ravi Sharma There are some different articles on this, but basically no.
** [[MikeBennett|Mike Bennett]]: Not unless you could write an ontology that defines what it is to be truly human
** Michael Robbins: Or rebuild the web from the bottom up to embed new frameworks for digital identity and content provenance/authenticity (which is what we need to commit to)
** [[AndreaWesterinen|Andrea Westerinen]]: Provenance also aids with detecting bots

* [[MichaelDeBellis|Michael DeBellis]]: Someone asked about agents in relation to Data Mesh (can't find the orig comment) IMO Data Mesh could be implemented in terms of Agents but typically isn't.

* [[RaviSharma|Ravi Sharma]]: Is there one kind or multiple kinds of connectivity in KG as well as in Ontologies?

* [[RaviSharma|Ravi Sharma]]: Container in the sense of domain overlaps especially overlapping vocabs as Venn diagrams?

* Anh: Why is it hard to build LLMs for other languages? Why can't it be replicated easily when translation work (e.g. Facebook, Google Translate, LinkedIn) has already done a somewhat good job?
** [[AndreaWesterinen|Andrea Westerinen]]: It is about the training data.
** [[BartGajderowicz|Bart Gajderowicz]]: [Anh], on longer text, translation looses many of the cultural nuances of a language, and looses context. Also, most training data is in English so most models are trained on English then translated.
** Anh: Thank you, @Andrea Westerinen & @Bart Gajderowicz. Does it meean it'd cost the same to build a new LLMs for a new language?
** [[AndreaWesterinen|Andrea Westerinen]]: I would believe so.
** [[BartGajderowicz|Bart Gajderowicz]]: The volume of English text for training is cheaper, it’s just the web. So I’d imagine finding enough text in your target language would be the biggest cost. Librarians are our friends here 🙂
** [[PennyAnderson|Penny Anderson]]: this context free is difficult when you start thinking of utility of UI
** Anh: Could you please elaborate more on utility of UI? @Penny Anderson
** [[PennyAnderson|Penny Anderson]]: I mean UIs tend to be process driven data-centric is not tightly bound to a particular process that is context free
** [[PennyAnderson|Penny Anderson]]: other than search ?
** Anh: I see. Thanks.
** [[PennyAnderson|Penny Anderson]]: Data providence, Ethical AI ?
** [[AndreaWesterinen|Andrea Westerinen]]: @Penny Anderson Much better declared and reasoned against in KGs.
** [[PennyAnderson|Penny Anderson]]: Zero trust in networks?

* [[RaviSharma|Ravi Sharma]]: Kurt when you talk of box, you are essentially stating in and out of scope items or is there a way of capturing the info outside the box and bringing it in?

* [[AndreaWesterinen|Andrea Westerinen]]: "What is valid for a graph? Shapes and data"

* [[MichaelDeBellis|Michael DeBellis]]: I don't agree that knowledge graphs are a form of data mesh. They are two related but different concepts. Data mesh is essentially an architecture for enterprise data definition and management. Knowledge graphs are a tool that can be used to implement a data mesh. Data mesh IMO brings the philosophy of microservices to data.
** Alan Morrison: Michael, re: microservices to data, should we be thinking of agents as messengers and KGs as the data resource>

* [[RaviSharma|Ravi Sharma]]: What kind of aggregates are these data shapes? Are these only valid for a class of data or can you mix data types in a shape aggregate?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi: in SHACL you define similar things as you do with OWL. E.g., does a property have to have exactly one value, the datatypes of a property, other constraints. The difference is OWL is used for reasoning over large knowledge graphs and uses the Open World Assumption. SHACL is for constraining data so it uses the Closed World Assumption. E.g., you can define ss_number as a property that must have exactly one value in either OWL or SHACL but in OWL you will almost never trigger an error if the restriction isn't satisfied due to OWA. With SHACL you will get error messages due to CWA

* [[ToddSchneider|Todd Schneider]]: LLM reasoning is probabilistic.

* Ayya Niyyanika Bhikkhuni: When weighting a concept is that some sort of credibility score?

* Michael Robbins: LangChains - I’m thinking about what role ConLangs could serve its intermediaries n revolutionizing language modes and NLP. Happy to have a follow-up discussion with anyone who is interested. https://en.wikipedia.org/wiki/Constructed_language

* [[RaviSharma|Ravi Sharma]]: What is the equivalent of Objects in LLM? What are these entities called and can same onto-entity be different in LLM context?
** [[MikeBennett|Mike Bennett]]: Words, I presume?

* [[RaviSharma|Ravi Sharma]]: Kurt thanks for including wonderful valuable background cultural images, these are inspiring. Are you also conveying the there is external (databased) and internal knowledge such as contemplation?

* [[RaviSharma|Ravi Sharma]]: Why are you limiting your examples to RDF why not MOF also?

* [[ToddSchneider|Todd Schneider]]: The important question is not mapping(s). It’s how can well constructed ontologies be used in the ingestion/training of LLMs.

* [[SusanneVejdemo|Sus Vejdemo]]: Question: what are good case studies of KGE and LLM integrations?
** [[AndreaWesterinen|Andrea Westerinen]]: I addressed some of this in the opening session. “Hybrid systems” include both where LLMs help ontologies (actually the Oct 25th session) and where ontologies help LLMs (Oct 4 and Nov 1 sessions).
** [[AndreaWesterinen|Andrea Westerinen]]: Also, Tony is highlighting a GREAT integration.

* [[MichaelDeBellis|Michael DeBellis]]: One question I have is what happens once you load a knowledge graph into an LLM? I know it can be done but once you load say a Turtle file into the LLM then what?
** [[BartGajderowicz|Bart Gajderowicz]]: Generally the LLM builds a small knowledge graph instance inside its memory and you can query it. Ask it to write SPARQL to get some instances, etc. I have not seen it used for large KGs, just small ones.

* Amit Jain: Will the recording be shared with attendees after the summit?
** [[MikeBennett|Mike Bennett]]: Yes, the recording will be uploaded to the session page when it is ready.

* Cedric Berger: Is there any metrics to measure that indeed KG combines with LLM are less hallucinating?
** [[AndreaWesterinen|Andrea Westerinen]]: I will try to provide these in the summary. I have read papers on this as well as blog posts.
** [[BartGajderowicz|Bart Gajderowicz]]: All the work I’ve come across relies on the knowledge graph to provide explicit knowledge. So if you can ground the LLM with a graph, you can verify if the answers the LLM provides are “facts” in the graph

* Kurt Cagle: Once you load in the ontology into a conversation, it will create (to some extent) an LLM conceptual space for that data. Also keep in mind that getting ALL of a knowledge graph via a RAG is usually not feasible.

* [[RaviSharma|Ravi Sharma]]: Tony great diagram for LLM + ontology to improve each other.why is ontology weak in capturing concepts Vs LLM?

* [[MichaelDeBellis|Michael DeBellis]]: One of the most interesting papers I've read is from Lawrence Berkeley Labs on using LLMs to extend an ontology.

* [[RaviSharma|Ravi Sharma]]: How would models such as in physics work with LLM and Ontology loop or cycle that you show, actually ontologies are conceptually richer than KGs alone?
** [[AndreaWesterinen|Andrea Westerinen]]: KGs are the data, ontologies are the concepts … So, it does not seem right to ask about one being richer.
** Michael Robbins: https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its-wolfram-superpowers/

* [[RaviSharma|Ravi Sharma]]: Tony the built-in uncertainty in LLMs gives it extra power to apply to real life probabilistic world?

* [[RaviSharma|Ravi Sharma]]: Tony LLMs and analog and ontology as Quantum? great way. thanks

* [[MikeBennett|Mike Bennett]]: Left brain v right brain is a great way of talking about ontology v LLM. Now you have to create a good corpus collosum.
** [[GaryBergCross|Gary Berg-Cross]]: A better model than left right hemispheres is by layers - old, mid brain (associative) and neo-cortex. They interconnect in many ways and some by the limbic system.
** [[BartGajderowicz|Bart Gajderowicz]]: I like the System 1 / System 2 analogy

* [[AndreaWesterinen|Andrea Westerinen]]: Tony characterized knowledge graphs as discrete and LLMs as continuous.

* Cedric Berger: Aren’t LLMs also kind of discrete as relying on vectors (arrays of numbers) of limited dimensions?
** [[BartGajderowicz|Bart Gajderowicz]]: LLMs are fundamentally probabilistic, not discrete. Some models are trained to provide discrete classifications, but that’s just at the output level. Internally they are probabilistic.
** [[AndreaWesterinen|Andrea Westerinen]]: The discreteness pertains to the encoding. But, does the probabilistic nature of LLMs make it more continuous? I am not sure.
** [[MikeBennett|Mike Bennett]]: The weightings are a number rather than a logical truth value as in KGs
** Kurt Cagle: Even with KGs, you can set up reifications that also set up Bayesians that are again more fuzzy (or at least more stochastic).
** [[MikeBennett|Mike Bennett]]: If you had an ontology with weightings instead of truth values and can train those values, you have a semantic network like a brain/mind.
** Kurt Cagle: It's where I think we're heading. People in the semantic space have known for years that knowledge is fuzzy / fractal, but getting there has always been the rub.
** [[MikeBennett|Mike Bennett]]: I roughed out an idea for this kind of semantic network application back in the 90s.

* Michael Robbins: A great article on vector embeddings: https://kdb.ai/learning-hub/fundamentals/vector-embeddings/ How can we use this for transparency and explainability? Give users confidence intervals (and other potential response options) along with responses?

* [[RaviSharma|Ravi Sharma]]: What are vectors equivalent to in LLM context?

* Harvey King: Are the embeddings stored across 3 layers?
** [[AndreaWesterinen|Andrea Westerinen]]: The embeddings are there, but simplified/reduced.

* Anh: Does KG consider the time stamp of the assertions/objects? Context of my question: could we use it to mark the originality of posts of similar contents to alert plagiarism.
** [[AndreaWesterinen|Andrea Westerinen]]: The KG CAN do this, if it is encoded.

* [[RaviSharma|Ravi Sharma]]: Has anyone done this a roundtrip quality check, learn from LLM and put it in ontologies and the other way around?
** Benoit Claise: In which context/use case?
** [[MichaelDeBellis|Michael DeBellis]]: That paper I posted earlier from Lawrence Berkeley Labs used LLM to extend an ontology but just went in one direction, expanding the ontology not changing the LLM.

* [[RaviSharma|Ravi Sharma]]: Can you do reasoning on same concept in both to differentiate their respective strengths and weakness?

* [[RaviSharma|Ravi Sharma]]: Tony your tree or chain of thought are great ways

* [[RaviSharma|Ravi Sharma]]: Tony and Kurt Can you address feature space Vs training set learning approaches?

* Liju Fan: Why are the relations in the ontology explicit?
** [[AndreaWesterinen|Andrea Westerinen]]: Relations are defined, and they can be inferred, but this is the essence of ontology. Ontologies are “open world” but do need relationships.
** [[MichaelDeBellis|Michael DeBellis]]: Because in the ontology you (usually manually) create the relations. In an LLM the relations are inferred by the ML algorithm and usually can't be manually changed
** Liju Fan: It seems there is a need to be able to rename LLM inferred relations for them to be human-understandable and practically useful.

* Kurt Cagle: Please note the similarity of Tony's slide with biological cells. Hmmm ...

* Cedric Berger: Has anyone asked AI to generate an image of a factual made of network base units?

* Michael Robbins: Agreed, Tony. And we’ve talked about this on LinkedIn. A constellation of domain-specific and ecosystem-based Community Knowledge Graphs and Language Models. #CKGs and #CLMs
** [[MikeBennett|Mike Bennett]]: TLO as mitochondria?
** Michael Robbins: Language is inseparable from culture and context

* Harvey King: Do KG's take a different nature when dealing with math?
** [[AndreaWesterinen|Andrea Westerinen]]: Not different, but with more rules?
** [[MichaelDeBellis|Michael DeBellis]]: Yes. An ontology uses explicit models like linear algebra. LLMs use linear algebra but computes answers based on examples, doesn't have a theoretical model of math (or other domains) as an ontology does.

* [[SusanneVejdemo|Sus Vejdemo]]: Thanks for a great talk! I love the conceptualization of embeddings as ontologies.

* [[MikeBennett|Mike Bennett]]: If you reify every edge you can give each one an analog value

* [[MarkUnderwood|Mark Underwood]]: For those interested in practical cybersec use cases (lots of chatty network data, some NLP), lot of narrow domain-specific emergent ontologies; e.g., Lambda / microservice mesh etc. mark.underwood@syf.com

* [[GaryBergCross|Gary Berg-Cross]]: I think the working memory graph is a useful early concept bring the LLMs and ontologies together but it is not so easy to capture what is the context for knowledge in this representation. I would guess this is a sub-set of the fluid knowledge of what human cognition employs. Much remains unconscious. But with research AI systems may make more of this explicit.

* [[RaviSharma|Ravi Sharma]]: My query is what is the relationship among reification, provenance and context history?

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Tony, “LLMs for compute” and “As much data into graph, then translating the graph paths to NL and adding to the LLM”

* [[GaryBergCross|Gary Berg-Cross]]: Reality may be atomistically discrete but at such a nano-level that continuous models make better predictions than discrete models that are orders of magnitude too gross rather than fine grained.

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Kurt: “Community Language Models - decentralized, federated, ad hoc network of information”

* [[GaryBergCross|Gary Berg-Cross]]: Models need to be more than federated. Because we center on semantic accuracy and relevance they need to be semantically harmonized.

== Resources ==
* [https://bit.ly/3S040lR Kurt Cagle Slides]
* [https://bit.ly/46A3EH2 Tony Seale Slides]
* [https://bit.ly/46DJyvo Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

AndreaWesterinen

2023-11-01T15:18:42Z

AndreaWesterinen:

= Andrea Westerinen =

Consultant and Researcher
'''Nine Points Solutions, LLC'''

email: arwesterinen [at] gmail.com
email: andreaw [at] ninepts.com

see: https://organizingknowledge.blogspot.com

see: https://hearing-all-voices.blogspot.com

see: http://www.ontoinsights.com

Andrea Westerinen is an independent software engineer and systems architect, and CTO of OntoInsights, LLC (http://ontoinsights.com). She specializes in ontology development and knowledge engineering, and has extensive software development experience. Ms. Westerinen has strong interests in semantic and linguistics technologies, and has worked in the computer industry since 1979, at places like Raytheon/BBN, Two Six Labs, SAIC, CA Technologies, Microsoft, Cisco, Intel and IBM. Her responsibilities have included researcher, strategist, program manager, personnel manager, software developer, ontologist and enthusiast, as needed. Ms. Westerinen has led and participated in many network and systems/storage management standards organizations, and held the positions of Technical Director for the Storage Networking Industry Association (SNIA) and Vice President of Technology for the Distributed Management Task Force (DMTF). She has a B.S. in Physics and Mathematics from Marquette University, and an M.S. in Computer Science from Nova Southeastern University. Ms. Westerinen currently holds 6 patents.

[[Category:Person]]

ConferenceCall 2023 11 01

2023-10-29T16:55:23Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 11 01

2023-10-24T16:09:21Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 11 01

2023-10-24T15:08:19Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 11 01

2023-10-24T15:08:06Z

AndreaWesterinen: /* Agenda */

OntologySummit2024

2023-10-24T15:06:01Z

AndreaWesterinen: /* Fall Series */

= [[OntologySummit2024|Ontology Summit 2024]] =

The [[OntologySummit|Ontology Summit]] is an annual series of events that involves the ontology community and communities related to each year's theme chosen for the summit. The Ontology Summit was started by Ontolog and NIST, and the program has been co-organized by Ontolog and NIST along with the co-sponsorship of other organizations that are supportive of the Summit goals and objectives.

== Purpose ==
As part of Ontolog’s general advocacy to bring ontology science and related engineering into the mainstream, we endeavor to facilitate discussion and knowledge sharing amongst stakeholders and interested parties relevant to the use of ontologies. The results will be synthesized and summarized in the form of the Ontology Summit 2024 Communiqué, with expanded supporting material provided on the web and in journal articles.

= Process and Deliverables =
Similar to our last seventeen summits, this [[OntologySummit2024|Ontology Summit 2024]] will consist of virtual discourse (over our archived mailing lists), virtual presentations and panel sessions as part of recorded video conference calls.
As in prior years the intent is to provide some synthesis of ideas and draft a communique summarizing major points.
This year will begin with a Fall Series in October and November; the main summit will begin in February.

Meetings are at Noon US/Canada Eastern Time on Wednesdays and last about an hour.

== Fall Series on Ontologies and Large Language Models: Related but Different ==

Fall Series Co-Chairs: Andrea Westerinen and Mike Bennett

{{:OntologySummit2024/Theme}}

= Schedule =
== Fall Series ==
* [[ConferenceCall_2023_10_04|4 October 2023]] ''Kickoff/Overview'', '''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''
* [[ConferenceCall_2023_10_11|11 October 2023]] ''Setting the stage'', '''[[DeborahMcGuinness|Deborah McGuinness]]'''
** Rennselaer Tetherless World Senior Constellation Chair
** Professor of Computer Science, Cognitive Science, and Industrial and Systems Engineering
** Expert in knowledge representation, reasoning languages and systems
* [[ConferenceCall_2023_10_18|18 October 2023]] ''A look across the industry, Part 1''
** '''Kurt Cagle''', Author of the [https://thecaglereport.com/ Cagle Report]
** '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/?originalSubdomain=uk LinkedIn])
* [[ConferenceCall_2023_10_25|25 October 2023]] ''A look across the industry, Part 2''
** '''Evren Sirin''', Stardog CTO and lead for their new [https://www.stardog.com/categories/voicebox/ Voicebox] offering
** '''Yuan He''', Key contributor to [https://github.com/KRR-Oxford/DeepOnto DeepOnto], a package for ontology engineering with deep learning
* [[ConferenceCall_2023_11_01|1 November 2023]] ''Demos of information extraction via hybrid systems''
** '''[[AndreaWesterinen|Andrea Westerinen]]''', Creator of [https://github.com/ontoinsights/deep_narrative_analysis Deep Narrative Analysis]
** '''[[PrasadYalamanchi|Prasad Yalamanchi]]''', [https://leadsemantics.com/ Lead Semantics] CTO
* [[ConferenceCall_2023_11_08|8 November 2023]] ''Broader thoughts''
** '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', Hybrid reasoning, the scope of knowledge, and what is beyond ontologies?
** '''[[JohnSowa|John Sowa]]''' and '''[[ArunMajumdar|Arun Majumdar]]''', LLMs, ontologies, and formal systems
* [[ConferenceCall_2023_11_15|15 November 2023]] Discussion and Synthesis, including questions for the full summit

= Resources =
* [[OntologySummit2024/ConferenceCallInformation|Conference Call Information]]
* [http://bit.ly/34DOmRV Ontology Summit YouTube Channel]

[[Category:OntologySummit]]
[[Category:OntologySummit2024]]

OntologySummit2024

2023-10-24T15:05:15Z

AndreaWesterinen: /* Fall Series */

= [[OntologySummit2024|Ontology Summit 2024]] =

The [[OntologySummit|Ontology Summit]] is an annual series of events that involves the ontology community and communities related to each year's theme chosen for the summit. The Ontology Summit was started by Ontolog and NIST, and the program has been co-organized by Ontolog and NIST along with the co-sponsorship of other organizations that are supportive of the Summit goals and objectives.

== Purpose ==
As part of Ontolog’s general advocacy to bring ontology science and related engineering into the mainstream, we endeavor to facilitate discussion and knowledge sharing amongst stakeholders and interested parties relevant to the use of ontologies. The results will be synthesized and summarized in the form of the Ontology Summit 2024 Communiqué, with expanded supporting material provided on the web and in journal articles.

= Process and Deliverables =
Similar to our last seventeen summits, this [[OntologySummit2024|Ontology Summit 2024]] will consist of virtual discourse (over our archived mailing lists), virtual presentations and panel sessions as part of recorded video conference calls.
As in prior years the intent is to provide some synthesis of ideas and draft a communique summarizing major points.
This year will begin with a Fall Series in October and November; the main summit will begin in February.

Meetings are at Noon US/Canada Eastern Time on Wednesdays and last about an hour.

== Fall Series on Ontologies and Large Language Models: Related but Different ==

Fall Series Co-Chairs: Andrea Westerinen and Mike Bennett

{{:OntologySummit2024/Theme}}

= Schedule =
== Fall Series ==
* [[ConferenceCall_2023_10_04|4 October 2023]] ''Kickoff/Overview'', '''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''
* [[ConferenceCall_2023_10_11|11 October 2023]] ''Setting the stage'', '''[[DeborahMcGuinness|Deborah McGuinness]]'''
** Rennselaer Tetherless World Senior Constellation Chair
** Professor of Computer Science, Cognitive Science, and Industrial and Systems Engineering
** Expert in knowledge representation, reasoning languages and systems
* [[ConferenceCall_2023_10_18|18 October 2023]] ''A look across the industry, Part 1''
** '''Kurt Cagle''', Author of the [https://thecaglereport.com/ Cagle Report]
** '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/?originalSubdomain=uk LinkedIn])
* [[ConferenceCall_2023_10_25|25 October 2023]] ''A look across the industry, Part 2''
** '''Evren Sirin''', Stardog CTO and lead for their new [https://www.stardog.com/categories/voicebox/ Voicebox] offering
** '''Yuan He''', Key contributor to [https://github.com/KRR-Oxford/DeepOnto DeepOnto], a package for ontology engineering with deep learning
* [[ConferenceCall_2023_11_01|1 November 2023]] ''Demos of information extraction via hybrid systems''
** '''[[AndreaWesterinen|Andrea Westerinen]]''', [https://github.com/ontoinsights/deep_narrative_analysis Deep Narrative Analysis]
** '''[[PrasadYalamanchi|Prasad Yalamanchi]]''', [https://leadsemantics.com/ Lead Semantics] CTO
* [[ConferenceCall_2023_11_08|8 November 2023]] ''Broader thoughts''
** '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', Hybrid reasoning, the scope of knowledge, and what is beyond ontologies?
** '''[[JohnSowa|John Sowa]]''' and '''[[ArunMajumdar|Arun Majumdar]]''', LLMs, ontologies, and formal systems
* [[ConferenceCall_2023_11_15|15 November 2023]] Discussion and Synthesis, including questions for the full summit

= Resources =
* [[OntologySummit2024/ConferenceCallInformation|Conference Call Information]]
* [http://bit.ly/34DOmRV Ontology Summit YouTube Channel]

[[Category:OntologySummit]]
[[Category:OntologySummit2024]]

OntologySummit2024

2023-10-24T15:04:48Z

AndreaWesterinen: /* Fall Series */

= [[OntologySummit2024|Ontology Summit 2024]] =

The [[OntologySummit|Ontology Summit]] is an annual series of events that involves the ontology community and communities related to each year's theme chosen for the summit. The Ontology Summit was started by Ontolog and NIST, and the program has been co-organized by Ontolog and NIST along with the co-sponsorship of other organizations that are supportive of the Summit goals and objectives.

== Purpose ==
As part of Ontolog’s general advocacy to bring ontology science and related engineering into the mainstream, we endeavor to facilitate discussion and knowledge sharing amongst stakeholders and interested parties relevant to the use of ontologies. The results will be synthesized and summarized in the form of the Ontology Summit 2024 Communiqué, with expanded supporting material provided on the web and in journal articles.

= Process and Deliverables =
Similar to our last seventeen summits, this [[OntologySummit2024|Ontology Summit 2024]] will consist of virtual discourse (over our archived mailing lists), virtual presentations and panel sessions as part of recorded video conference calls.
As in prior years the intent is to provide some synthesis of ideas and draft a communique summarizing major points.
This year will begin with a Fall Series in October and November; the main summit will begin in February.

Meetings are at Noon US/Canada Eastern Time on Wednesdays and last about an hour.

== Fall Series on Ontologies and Large Language Models: Related but Different ==

Fall Series Co-Chairs: Andrea Westerinen and Mike Bennett

{{:OntologySummit2024/Theme}}

= Schedule =
== Fall Series ==
* [[ConferenceCall_2023_10_04|4 October 2023]] ''Kickoff/Overview'', '''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''
* [[ConferenceCall_2023_10_11|11 October 2023]] ''Setting the stage'', '''[[DeborahMcGuinness|Deborah McGuinness]]'''
** Rennselaer Tetherless World Senior Constellation Chair
** Professor of Computer Science, Cognitive Science, and Industrial and Systems Engineering
** Expert in knowledge representation, reasoning languages and systems
* [[ConferenceCall_2023_10_18|18 October 2023]] ''A look across the industry, Part 1''
** '''Kurt Cagle''', Author of the [https://thecaglereport.com/ Cagle Report]
** '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/?originalSubdomain=uk LinkedIn])
* [[ConferenceCall_2023_10_25|25 October 2023]] ''A look across the industry, Part 2''
** '''Evren Sirin''', Stardog CTO and lead for their new [https://www.stardog.com/categories/voicebox/ Voicebox] offering
** '''Yuan He''', Key contributor to [https://github.com/KRR-Oxford/DeepOnto DeepOnto], a package for ontology engineering with deep learning
* [[ConferenceCall_2023_11_01|1 November 2023]] ''Demos of information extraction via hybrid systems''
** '''[[AndreaWesterinen|Andrea Westerinen]]''', [https://github.com/ontoinsights/deep_narrative_analysis DNA, Deep Narrative Analysis]
** '''[[PrasadYalamanchi|Prasad Yalamanchi]]''', [https://leadsemantics.com/ Lead Semantics] CTO
* [[ConferenceCall_2023_11_08|8 November 2023]] ''Broader thoughts''
** '''[[AnatolyLevenchuk|Anatoly Levenchuk]]''', Hybrid reasoning, the scope of knowledge, and what is beyond ontologies?
** '''[[JohnSowa|John Sowa]]''' and '''[[ArunMajumdar|Arun Majumdar]]''', LLMs, ontologies, and formal systems
* [[ConferenceCall_2023_11_15|15 November 2023]] Discussion and Synthesis, including questions for the full summit

= Resources =
* [[OntologySummit2024/ConferenceCallInformation|Conference Call Information]]
* [http://bit.ly/34DOmRV Ontology Summit YouTube Channel]

[[Category:OntologySummit]]
[[Category:OntologySummit2024]]

ConferenceCall 2023 11 01

2023-10-23T14:06:02Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 10 18

2023-10-23T03:17:15Z

AndreaWesterinen: /* Agenda */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 1]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::18 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Kurt Cagle''', Author of [https://thecaglereport.com/ The Cagle Report]
** '''Title:''' Complementary Thinking: Language Models, Ontologies and Knowledge Graphs
** '''Abstract:''' With the advent of Retrieval Augmented Generators (RAGs), a more or less standardized workflow has become available for integrating large language models such as ChatGPT with knowledge graphs. This in turn has raised the question about the nature of ontologies associated with LLMs and how knowledge graphs can be structured and queried to make integrated data access possible between the two types of systems. In this talk, Editor and AI Explorer Kurt Cagle of The Cagle Report looks at this process and discusses how they affect both knowledge portals and ontology design.
** [https://bit.ly/3S040lR Slides]
* '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/ LinkedIn])
** '''Title:''' How Ontologies Can Unlock the Potential of Large Language Models for Business
** '''Abstract:''' LLMs have remarkable capabilities; they can craft letters, analyze data, orchestrate workflows, generate code, and much more. Companies such as Google, Apple, Amazon, Meta, and Microsoft are all investing heavily in this technology. Everything indicates that LLMs have enormous disruptive potential. However, there is a problem: they can hallucinate, and for any serious business, that is a deal-breaker. This is where ontologies can come in. In combination with Knowledge Graphs, they can place guardrails around the LLMs, thus allowing organizations to harness the capabilities of LLMs within the framework of a safely controlled ontological structure.
** [https://bit.ly/46A3EH2 Slides]

[https://bit.ly/46DJyvo Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 18 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=18&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
80+ participants

* Kurt Cagle
* Tony Seale
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* Ayya Niyyanika Bhikkhuni
* Anthony Alcaraz
* Alan Morrison
* [[BartGajderowicz|Bart Gajderowicz]]
* Chris Day
* Andrew McCaffrey
* [[GaryBergCross|Gary Berg-Cross]]
* [[JohnSowa|John Sowa]]
* [[MarkUnderwood|Mark Underwood (IS Innovation)]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[SusanneVejdemo|Sus Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* Yishan Liu
* James Logan
* [[PennyAnderson|Penny Anderson]]
* Michael Robbins
* [[MichaelDeBellis|Michael DeBellis]]
* Amit Jain
* Cedric Berger
* Andreas Lothe Opdahl
* Harvey King
* Benoit Claise
* Liju Fan
* Larry Swanson
* Justin Lewis

== Discussion ==
* Anthony's OtterPilot: Hi, I'm an AI assistant helping Anthony Alcaraz take notes for this meeting. Follow along the transcript here: https://otter.ai/u/WLIaj2w-OmOEoVVCP5gURhhZHaY?utm_source=va_chat_link_2  You'll also be able to see screenshots of key moments, add highlights, comments, or action items to anything being said, and get an automatic summary after the meeting.
** [[MikeBennett|Mike Bennett]]: Dang! Put on a session on AI and the AIs start showing up!

* [[AndreaWesterinen|Andrea Westerinen]]: I like the quote, “Ontologies are the shapes of information and knowledge”

* [[ToddSchneider|Todd Schneider]]: What is a ‘shape of information’?
** [[AndreaWesterinen|Andrea Westerinen]]: Roles, constraints, relationships for a domain; a local representation

* [[RaviSharma|Ravi Sharma]]: Can LLMs and/or ontologies be used to detect AI artifacts?
** [[AndreaWesterinen|Andrea Westerinen]]: @Ravi Sharma There are some different articles on this, but basically no.
** [[MikeBennett|Mike Bennett]]: Not unless you could write an ontology that defines what it is to be truly human
** Michael Robbins: Or rebuild the web from the bottom up to embed new frameworks for digital identity and content provenance/authenticity (which is what we need to commit to)
** [[AndreaWesterinen|Andrea Westerinen]]: Provenance also aids with detecting bots

* [[MichaelDeBellis|Michael DeBellis]]: Someone asked about agents in relation to Data Mesh (can't find the orig comment) IMO Data Mesh could be implemented in terms of Agents but typically isn't.

* [[RaviSharma|Ravi Sharma]]: Is there one kind or multiple kinds of connectivity in KG as well as in Ontologies?

* [[RaviSharma|Ravi Sharma]]: Container in the sense of domain overlaps especially overlapping vocabs as Venn diagrams?

* Anh: Why is it hard to build LLMs for other languages? Why can't it be replicated easily when translation work (e.g. Facebook, Google Translate, LinkedIn) has already done a somewhat good job?
** [[AndreaWesterinen|Andrea Westerinen]]: It is about the training data.
** [[BartGajderowicz|Bart Gajderowicz]]: [Anh], on longer text, translation looses many of the cultural nuances of a language, and looses context. Also, most training data is in English so most models are trained on English then translated.
** Anh: Thank you, @Andrea Westerinen & @Bart Gajderowicz. Does it meean it'd cost the same to build a new LLMs for a new language?
** [[AndreaWesterinen|Andrea Westerinen]]: I would believe so.
** [[BartGajderowicz|Bart Gajderowicz]]: The volume of English text for training is cheaper, it’s just the web. So I’d imagine finding enough text in your target language would be the biggest cost. Librarians are our friends here 🙂
** [[PennyAnderson|Penny Anderson]]: this context free is difficult when you start thinking of utility of UI
** Anh: Could you please elaborate more on utility of UI? @Penny Anderson
** [[PennyAnderson|Penny Anderson]]: I mean UIs tend to be process driven data-centric is not tightly bound to a particular process that is context free
** [[PennyAnderson|Penny Anderson]]: other than search ?
** Anh: I see. Thanks.
** [[PennyAnderson|Penny Anderson]]: Data providence, Ethical AI ?
** [[AndreaWesterinen|Andrea Westerinen]]: @Penny Anderson Much better declared and reasoned against in KGs.
** [[PennyAnderson|Penny Anderson]]: Zero trust in networks?

* [[RaviSharma|Ravi Sharma]]: Kurt when you talk of box, you are essentially stating in and out of scope items or is there a way of capturing the info outside the box and bringing it in?

* [[MichaelDeBellis|Michael DeBellis]]: I don't agree that knowledge graphs are a form of data mesh. They are two related but different concepts. Data mesh is essentially an architecture for enterprise data definition and management. Knowledge graphs are a tool that can be used to implement a data mesh. Data mesh IMO brings the philosophy of microservices to data.
** Alan Morrison: Michael, re: microservices to data, should we be thinking of agents as messengers and KGs as the data resource>

* [[RaviSharma|Ravi Sharma]]: What kind of aggregates are these data shapes? Are these only valid for a class of data or can you mix data types in a shape aggregate?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi: in SHACL you define similar things as you do with OWL. E.g., does a property have to have exactly one value, the datatypes of a property, other constraints. The difference is OWL is used for reasoning over large knowledge graphs and uses the Open World Assumption. SHACL is for constraining data so it uses the Closed World Assumption. E.g., you can define ss_number as a property that must have exactly one value in either OWL or SHACL but in OWL you will almost never trigger an error if the restriction isn't satisfied due to OWA. With SHACL you will get error messages due to CWA

* [[ToddSchneider|Todd Schneider]]: LLM reasoning is probabilistic.

* Ayya Niyyanika Bhikkhuni: When weighting a concept is that some sort of credibility score?

* Michael Robbins: LangChains - I’m thinking about what role ConLangs could serve its intermediaries n revolutionizing language modes and NLP. Happy to have a follow-up discussion with anyone who is interested. https://en.wikipedia.org/wiki/Constructed_language

* [[RaviSharma|Ravi Sharma]]: What is the equivalent of Objects in LLM? What are these entities called and can same onto-entity be different in LLM context?
** [[MikeBennett|Mike Bennett]]: Words, I presume?

* [[RaviSharma|Ravi Sharma]]: Kurt thanks for including wonderful valuable background cultural images, these are inspiring. Are you also conveying the there is external (databased) and internal knowledge such as contemplation?

* [[RaviSharma|Ravi Sharma]]: Why are you limiting your examples to RDF why not MOF also?

* [[ToddSchneider|Todd Schneider]]: The important question is not mapping(s). It’s how can well constructed ontologies be used in the ingestion/training of LLMs.

* [[SusanneVejdemo|Sus Vejdemo]]: Question: what are good case studies of KGE and LLM integrations?
** [[AndreaWesterinen|Andrea Westerinen]]: I addressed some of this in the opening session. “Hybrid systems” include both where LLMs help ontologies (actually the Oct 25th session) and where ontologies help LLMs (Oct 4 and Nov 1 sessions).
** [[AndreaWesterinen|Andrea Westerinen]]: Also, Tony is highlighting a GREAT integration.

* [[MichaelDeBellis|Michael DeBellis]]: One question I have is what happens once you load a knowledge graph into an LLM? I know it can be done but once you load say a Turtle file into the LLM then what?
** [[BartGajderowicz|Bart Gajderowicz]]: Generally the LLM builds a small knowledge graph instance inside its memory and you can query it. Ask it to write SPARQL to get some instances, etc. I have not seen it used for large KGs, just small ones.

* Amit Jain: Will the recording be shared with attendees after the summit?
** [[MikeBennett|Mike Bennett]]: Yes, the recording will be uploaded to the session page when it is ready.

* Cedric Berger: Is there any metrics to measure that indeed KG combines with LLM are less hallucinating?
** [[AndreaWesterinen|Andrea Westerinen]]: I will try to provide these in the summary. I have read papers on this as well as blog posts.
** [[BartGajderowicz|Bart Gajderowicz]]: All the work I’ve come across relies on the knowledge graph to provide explicit knowledge. So if you can ground the LLM with a graph, you can verify if the answers the LLM provides are “facts” in the graph

* Kurt Cagle: Once you load in the ontology into a conversation, it will create (to some extent) an LLM conceptual space for that data. Also keep in mind that getting ALL of a knowledge graph via a RAG is usually not feasible.

* [[RaviSharma|Ravi Sharma]]: Tony great diagram for LLM ontology lop to improve each other.why is ontology weak in capturing concepts Vs LLM?

* [[MichaelDeBellis|Michael DeBellis]]: One of the most interesting papers I've read is from Lawrence Berkeley Labs on using LLMs to extend an ontology.

* [[RaviSharma|Ravi Sharma]]: How would models such as in physics work with LLM and Ontology loop or cycle that you show, actually ontologies are conceptually richer than KGs alone?
** [[AndreaWesterinen|Andrea Westerinen]]: KGs are the data, ontologies are the concepts … So, it does not seem right to ask about one being richer.
** Michael Robbins: https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its-wolfram-superpowers/

* [[RaviSharma|Ravi Sharma]]: Tony the built-in ubcertainty in LLMs gives it extra power to apply to real life probabilistic world?

* [[RaviSharma|Ravi Sharma]]: Tony LLMs and analog and ontology as Quantum? great way. thanks

* [[MikeBennett|Mike Bennett]]: Left brain v right brain is a great way of talking about ontology v LLM. Now you have to create a good corpus collosum.
** [[GaryBergCross|Gary Berg-Cross]]: A better model than left right hemispheres is by layers - old, mid brain (associative) and neo-cortex. They interconnect in many ways and some by the limbic system.
** [[BartGajderowicz|Bart Gajderowicz]]: I like the System 1 / System 2 analogy

* Cedric Berger: Aren’t LLMs also kind of discrete as relying on vectors (arrays of numbers) of limited dimensions?
** [[BartGajderowicz|Bart Gajderowicz]]: LLMs are fundamentally probabilistic, not discrete. Some models are trained to provide discrete classifications, but that’s just at the output level. Internally they are probabilistic.
** [[AndreaWesterinen|Andrea Westerinen]]: The discreteness pertains to the encoding. But, does the probabilistic nature of LLMs make it more continuous? I am not sure.
** [[MikeBennett|Mike Bennett]]: The weightings are a number rather than a logical truth value as in KGs
** Kurt Cagle: Even with KGs, you can set up reifications that also set up Bayesians that are again more fuzzy (or at least more stochastic).
** [[MikeBennett|Mike Bennett]]: If you had an ontology with weightings instead of truth values and can train those values, you have a semantic network like a brain/mind.
** Kurt Cagle: It's where I think we're heading. People in the semantic space have known for years that knowledge is fuzzy / fractal, but getting there has always been the rub.
** [[MikeBennett|Mike Bennett]]: I roughed out an idea for this kind of semantic network application back in the 90s.

* Michael Robbins: A great article on vector embeddings: https://kdb.ai/learning-hub/fundamentals/vector-embeddings/ How can we use this for transparency and explainability? Give users confidence intervals (and other potential response options) along with responses?

* [[RaviSharma|Ravi Sharma]]: What are vectors equivalent to in LLM context?

* Harvey King: Are the embeddings stored across 3 layers?
** [[AndreaWesterinen|Andrea Westerinen]]: The embeddings are there, but simplified/reduced.

* Anh: Does KG consider the time stamp of the assertions/objects? Context of my question: could we use it to mark the originality of posts of similar contents to alert plagiarism.
** [[AndreaWesterinen|Andrea Westerinen]]: The KG CAN do this, if it is encoded.

* [[RaviSharma|Ravi Sharma]]: Has anyone done this a roundtrip quality check, learn from LLM and put it in ontologies and the other way around?
** Benoit Claise: In which context/use case?
** [[MichaelDeBellis|Michael DeBellis]]: That paper I posted earlier from Lawrence Berkeley Labs used LLM to extend an ontology but just went in one direction, expanding the ontology not changing the LLM.

* [[RaviSharma|Ravi Sharma]]: Can you do reasoning on same concept in both to differentiate their respective strengths and weakness?

* [[RaviSharma|Ravi Sharma]]: Tony your tree or chain of thought are great ways

* [[RaviSharma|Ravi Sharma]]: Tony and Kurt Can you address feature space Vs training set learning approaches?

* Liju Fan: Why are the relations in the ontology explicit?
** [[AndreaWesterinen|Andrea Westerinen]]: Relations are defined, and they can be inferred, but this is the essence of ontology. Ontologies are “open world” but do need relationships.
** [[MichaelDeBellis|Michael DeBellis]]: Because in the ontology you (usually manually) create the relations. In an LLM the relations are inferred by the ML algorithm and usually can't be manually changed
** Liju Fan: It seems there is a need to be able to rename LLM inferred relations for them to be human-understandable and practically useful.

* Kurt Cagle: Please note the similarity of Tony's slide with biological cells. Hmmm ...

* Cedric Berger: Has anyone asked AI to generate an image of a factual made of network base units?

* Michael Robbins: Agreed, Tony. And we’ve talked about this on LinkedIn. A constellation of domain-specific and ecosystem-based Community Knowledge Graphs and Language Models. #CKGs and #CLMs
** [[MikeBennett|Mike Bennett]]: TLO as mitochondria?
** Michael Robbins: Language is inseparable from culture and context

* Harvey King: Do KG's take a different nature when dealing with math?
** [[AndreaWesterinen|Andrea Westerinen]]: Not different, but with more rules?
** [[MichaelDeBellis|Michael DeBellis]]: Yes. An ontology uses explicit models like linear algebra. LLMs use linear algebra but computes answers based on examples, doesn't have a theoretical model of math (or other domains) as an ontology does.

* [[SusanneVejdemo|Sus Vejdemo]]: Thanks for a great talk! I love the conceptualization of embeddings as ontologies.

* [[MikeBennett|Mike Bennett]]: If you reify every edge you can give each one an analog value

* [[MarkUnderwood|Mark Underwood]]: For those interested in practical cybersec use cases (lots of chatty network data, some NLP), lot of narrow domain-specific emergent ontologies; e.g., Lambda / microservice mesh etc. mark.underwood@syf.com

* [[GaryBergCross|Gary Berg-Cross]]: I think the working memory graph is a useful early concept bring the LLMs and ontologies together but it is not so easy to capture what is the context for knowledge in this representation. I would guess this is a sub-set of the fluid knowledge of what human cognition employs. Much remains unconscious. But with research AI systems may make more of this explicit.

* [[RaviSharma|Ravi Sharma]]: My query is what is the relationship among reification, provenance and context history?

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Tony, “LLMs for compute” and “As much data into graph, then translating the graph paths to NL and adding to the LLM”

* [[GaryBergCross|Gary Berg-Cross]]: Reality may be atomistically discrete but at such a nano-level that continuous models make better predictions than discrete models that are orders of magnitude too gross rather than fine grained.

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Kurt: “Community Language Models - decentralized, federated, ad hoc network of information”

* [[GaryBergCross|Gary Berg-Cross]]: Models need to be more than federated. Because we center on semantic accuracy and relevance they need to be semantically harmonized.

== Resources ==
* [https://bit.ly/46DJyvo Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 18

2023-10-23T03:02:26Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 1]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::18 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Kurt Cagle''', Author of [https://thecaglereport.com/ The Cagle Report]
** '''Title:''' Complementary Thinking: Language Models, Ontologies and Knowledge Graphs
** '''Abstract:''' With the advent of Retrieval Augmented Generators (RAGs), a more or less standardized workflow has become available for integrating large language models such as ChatGPT with knowledge graphs. This in turn has raised the question about the nature of ontologies associated with LLMs and how knowledge graphs can be structured and queried to make integrated data access possible between the two types of systems. In this talk, Editor and AI Explorer Kurt Cagle of The Cagle Report looks at this process and discusses how they affect both knowledge portals and ontology design.
** [https://bit.ly/3S040lR Slides]
* '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/ LinkedIn])
** '''Title:''' How Ontologies Can Unlock the Potential of Large Language Models for Business
** '''Abstract:''' LLMs have remarkable capabilities; they can craft letters, analyze data, orchestrate workflows, generate code, and much more. Companies such as Google, Apple, Amazon, Meta, and Microsoft are all investing heavily in this technology. Everything indicates that LLMs have enormous disruptive potential. However, there is a problem: they can hallucinate, and for any serious business, that is a deal-breaker. This is where ontologies can come in. In combination with Knowledge Graphs, they can place guardrails around the LLMs, thus allowing organizations to harness the capabilities of LLMs within the framework of a safely controlled ontological structure.

[https://bit.ly/46A3EH2 Slides]

[https://bit.ly/46DJyvo Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 18 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=18&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
80+ participants

* Kurt Cagle
* Tony Seale
* [[KenBaclawski|Ken Baclawski]]
* [[AndreaWesterinen|Andrea Westerinen]]
* Ayya Niyyanika Bhikkhuni
* Anthony Alcaraz
* Alan Morrison
* [[BartGajderowicz|Bart Gajderowicz]]
* Chris Day
* Andrew McCaffrey
* [[GaryBergCross|Gary Berg-Cross]]
* [[JohnSowa|John Sowa]]
* [[MarkUnderwood|Mark Underwood (IS Innovation)]]
* [[MikeBennett|Mike Bennett]]
* [[RaviSharma|Ravi Sharma]]
* [[SusanneVejdemo|Sus Vejdemo]]
* [[ToddSchneider|Todd Schneider]]
* Yishan Liu
* James Logan
* [[PennyAnderson|Penny Anderson]]
* Michael Robbins
* [[MichaelDeBellis|Michael DeBellis]]
* Amit Jain
* Cedric Berger
* Andreas Lothe Opdahl
* Harvey King
* Benoit Claise
* Liju Fan
* Larry Swanson
* Justin Lewis

== Discussion ==
* Anthony's OtterPilot: Hi, I'm an AI assistant helping Anthony Alcaraz take notes for this meeting. Follow along the transcript here: https://otter.ai/u/WLIaj2w-OmOEoVVCP5gURhhZHaY?utm_source=va_chat_link_2  You'll also be able to see screenshots of key moments, add highlights, comments, or action items to anything being said, and get an automatic summary after the meeting.
** [[MikeBennett|Mike Bennett]]: Dang! Put on a session on AI and the AIs start showing up!

* [[AndreaWesterinen|Andrea Westerinen]]: I like the quote, “Ontologies are the shapes of information and knowledge”

* [[ToddSchneider|Todd Schneider]]: What is a ‘shape of information’?
** [[AndreaWesterinen|Andrea Westerinen]]: Roles, constraints, relationships for a domain; a local representation

* [[RaviSharma|Ravi Sharma]]: Can LLMs and/or ontologies be used to detect AI artifacts?
** [[AndreaWesterinen|Andrea Westerinen]]: @Ravi Sharma There are some different articles on this, but basically no.
** [[MikeBennett|Mike Bennett]]: Not unless you could write an ontology that defines what it is to be truly human
** Michael Robbins: Or rebuild the web from the bottom up to embed new frameworks for digital identity and content provenance/authenticity (which is what we need to commit to)
** [[AndreaWesterinen|Andrea Westerinen]]: Provenance also aids with detecting bots

* [[MichaelDeBellis|Michael DeBellis]]: Someone asked about agents in relation to Data Mesh (can't find the orig comment) IMO Data Mesh could be implemented in terms of Agents but typically isn't.

* [[RaviSharma|Ravi Sharma]]: Is there one kind or multiple kinds of connectivity in KG as well as in Ontologies?

* [[RaviSharma|Ravi Sharma]]: Container in the sense of domain overlaps especially overlapping vocabs as Venn diagrams?

* Anh: Why is it hard to build LLMs for other languages? Why can't it be replicated easily when translation work (e.g. Facebook, Google Translate, LinkedIn) has already done a somewhat good job?
** [[AndreaWesterinen|Andrea Westerinen]]: It is about the training data.
** [[BartGajderowicz|Bart Gajderowicz]]: [Anh], on longer text, translation looses many of the cultural nuances of a language, and looses context. Also, most training data is in English so most models are trained on English then translated.
** Anh: Thank you, @Andrea Westerinen & @Bart Gajderowicz. Does it meean it'd cost the same to build a new LLMs for a new language?
** [[AndreaWesterinen|Andrea Westerinen]]: I would believe so.
** [[BartGajderowicz|Bart Gajderowicz]]: The volume of English text for training is cheaper, it’s just the web. So I’d imagine finding enough text in your target language would be the biggest cost. Librarians are our friends here 🙂
** [[PennyAnderson|Penny Anderson]]: this context free is difficult when you start thinking of utility of UI
** Anh: Could you please elaborate more on utility of UI? @Penny Anderson
** [[PennyAnderson|Penny Anderson]]: I mean UIs tend to be process driven data-centric is not tightly bound to a particular process that is context free
** [[PennyAnderson|Penny Anderson]]: other than search ?
** Anh: I see. Thanks.
** [[PennyAnderson|Penny Anderson]]: Data providence, Ethical AI ?
** [[AndreaWesterinen|Andrea Westerinen]]: @Penny Anderson Much better declared and reasoned against in KGs.
** [[PennyAnderson|Penny Anderson]]: Zero trust in networks?

* [[RaviSharma|Ravi Sharma]]: Kurt when you talk of box, you are essentially stating in and out of scope items or is there a way of capturing the info outside the box and bringing it in?

* [[MichaelDeBellis|Michael DeBellis]]: I don't agree that knowledge graphs are a form of data mesh. They are two related but different concepts. Data mesh is essentially an architecture for enterprise data definition and management. Knowledge graphs are a tool that can be used to implement a data mesh. Data mesh IMO brings the philosophy of microservices to data.
** Alan Morrison: Michael, re: microservices to data, should we be thinking of agents as messengers and KGs as the data resource>

* [[RaviSharma|Ravi Sharma]]: What kind of aggregates are these data shapes? Are these only valid for a class of data or can you mix data types in a shape aggregate?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi: in SHACL you define similar things as you do with OWL. E.g., does a property have to have exactly one value, the datatypes of a property, other constraints. The difference is OWL is used for reasoning over large knowledge graphs and uses the Open World Assumption. SHACL is for constraining data so it uses the Closed World Assumption. E.g., you can define ss_number as a property that must have exactly one value in either OWL or SHACL but in OWL you will almost never trigger an error if the restriction isn't satisfied due to OWA. With SHACL you will get error messages due to CWA

* [[ToddSchneider|Todd Schneider]]: LLM reasoning is probabilistic.

* Ayya Niyyanika Bhikkhuni: When weighting a concept is that some sort of credibility score?

* Michael Robbins: LangChains - I’m thinking about what role ConLangs could serve its intermediaries n revolutionizing language modes and NLP. Happy to have a follow-up discussion with anyone who is interested. https://en.wikipedia.org/wiki/Constructed_language

* [[RaviSharma|Ravi Sharma]]: What is the equivalent of Objects in LLM? What are these entities called and can same onto-entity be different in LLM context?
** [[MikeBennett|Mike Bennett]]: Words, I presume?

* [[RaviSharma|Ravi Sharma]]: Kurt thanks for including wonderful valuable background cultural images, these are inspiring. Are you also conveying the there is external (databased) and internal knowledge such as contemplation?

* [[RaviSharma|Ravi Sharma]]: Why are you limiting your examples to RDF why not MOF also?

* [[ToddSchneider|Todd Schneider]]: The important question is not mapping(s). It’s how can well constructed ontologies be used in the ingestion/training of LLMs.

* [[SusanneVejdemo|Sus Vejdemo]]: Question: what are good case studies of KGE and LLM integrations?
** [[AndreaWesterinen|Andrea Westerinen]]: I addressed some of this in the opening session. “Hybrid systems” include both where LLMs help ontologies (actually the Oct 25th session) and where ontologies help LLMs (Oct 4 and Nov 1 sessions).
** [[AndreaWesterinen|Andrea Westerinen]]: Also, Tony is highlighting a GREAT integration.

* [[MichaelDeBellis|Michael DeBellis]]: One question I have is what happens once you load a knowledge graph into an LLM? I know it can be done but once you load say a Turtle file into the LLM then what?
** [[BartGajderowicz|Bart Gajderowicz]]: Generally the LLM builds a small knowledge graph instance inside its memory and you can query it. Ask it to write SPARQL to get some instances, etc. I have not seen it used for large KGs, just small ones.

* Amit Jain: Will the recording be shared with attendees after the summit?
** [[MikeBennett|Mike Bennett]]: Yes, the recording will be uploaded to the session page when it is ready.

* Cedric Berger: Is there any metrics to measure that indeed KG combines with LLM are less hallucinating?
** [[AndreaWesterinen|Andrea Westerinen]]: I will try to provide these in the summary. I have read papers on this as well as blog posts.
** [[BartGajderowicz|Bart Gajderowicz]]: All the work I’ve come across relies on the knowledge graph to provide explicit knowledge. So if you can ground the LLM with a graph, you can verify if the answers the LLM provides are “facts” in the graph

* Kurt Cagle: Once you load in the ontology into a conversation, it will create (to some extent) an LLM conceptual space for that data. Also keep in mind that getting ALL of a knowledge graph via a RAG is usually not feasible.

* [[RaviSharma|Ravi Sharma]]: Tony great diagram for LLM ontology lop to improve each other.why is ontology weak in capturing concepts Vs LLM?

* [[MichaelDeBellis|Michael DeBellis]]: One of the most interesting papers I've read is from Lawrence Berkeley Labs on using LLMs to extend an ontology.

* [[RaviSharma|Ravi Sharma]]: How would models such as in physics work with LLM and Ontology loop or cycle that you show, actually ontologies are conceptually richer than KGs alone?
** [[AndreaWesterinen|Andrea Westerinen]]: KGs are the data, ontologies are the concepts … So, it does not seem right to ask about one being richer.
** Michael Robbins: https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its-wolfram-superpowers/

* [[RaviSharma|Ravi Sharma]]: Tony the built-in ubcertainty in LLMs gives it extra power to apply to real life probabilistic world?

* [[RaviSharma|Ravi Sharma]]: Tony LLMs and analog and ontology as Quantum? great way. thanks

* [[MikeBennett|Mike Bennett]]: Left brain v right brain is a great way of talking about ontology v LLM. Now you have to create a good corpus collosum.
** [[GaryBergCross|Gary Berg-Cross]]: A better model than left right hemispheres is by layers - old, mid brain (associative) and neo-cortex. They interconnect in many ways and some by the limbic system.
** [[BartGajderowicz|Bart Gajderowicz]]: I like the System 1 / System 2 analogy

* Cedric Berger: Aren’t LLMs also kind of discrete as relying on vectors (arrays of numbers) of limited dimensions?
** [[BartGajderowicz|Bart Gajderowicz]]: LLMs are fundamentally probabilistic, not discrete. Some models are trained to provide discrete classifications, but that’s just at the output level. Internally they are probabilistic.
** [[AndreaWesterinen|Andrea Westerinen]]: The discreteness pertains to the encoding. But, does the probabilistic nature of LLMs make it more continuous? I am not sure.
** [[MikeBennett|Mike Bennett]]: The weightings are a number rather than a logical truth value as in KGs
** Kurt Cagle: Even with KGs, you can set up reifications that also set up Bayesians that are again more fuzzy (or at least more stochastic).
** [[MikeBennett|Mike Bennett]]: If you had an ontology with weightings instead of truth values and can train those values, you have a semantic network like a brain/mind.
** Kurt Cagle: It's where I think we're heading. People in the semantic space have known for years that knowledge is fuzzy / fractal, but getting there has always been the rub.
** [[MikeBennett|Mike Bennett]]: I roughed out an idea for this kind of semantic network application back in the 90s.

* Michael Robbins: A great article on vector embeddings: https://kdb.ai/learning-hub/fundamentals/vector-embeddings/ How can we use this for transparency and explainability? Give users confidence intervals (and other potential response options) along with responses?

* [[RaviSharma|Ravi Sharma]]: What are vectors equivalent to in LLM context?

* Harvey King: Are the embeddings stored across 3 layers?
** [[AndreaWesterinen|Andrea Westerinen]]: The embeddings are there, but simplified/reduced.

* Anh: Does KG consider the time stamp of the assertions/objects? Context of my question: could we use it to mark the originality of posts of similar contents to alert plagiarism.
** [[AndreaWesterinen|Andrea Westerinen]]: The KG CAN do this, if it is encoded.

* [[RaviSharma|Ravi Sharma]]: Has anyone done this a roundtrip quality check, learn from LLM and put it in ontologies and the other way around?
** Benoit Claise: In which context/use case?
** [[MichaelDeBellis|Michael DeBellis]]: That paper I posted earlier from Lawrence Berkeley Labs used LLM to extend an ontology but just went in one direction, expanding the ontology not changing the LLM.

* [[RaviSharma|Ravi Sharma]]: Can you do reasoning on same concept in both to differentiate their respective strengths and weakness?

* [[RaviSharma|Ravi Sharma]]: Tony your tree or chain of thought are great ways

* [[RaviSharma|Ravi Sharma]]: Tony and Kurt Can you address feature space Vs training set learning approaches?

* Liju Fan: Why are the relations in the ontology explicit?
** [[AndreaWesterinen|Andrea Westerinen]]: Relations are defined, and they can be inferred, but this is the essence of ontology. Ontologies are “open world” but do need relationships.
** [[MichaelDeBellis|Michael DeBellis]]: Because in the ontology you (usually manually) create the relations. In an LLM the relations are inferred by the ML algorithm and usually can't be manually changed
** Liju Fan: It seems there is a need to be able to rename LLM inferred relations for them to be human-understandable and practically useful.

* Kurt Cagle: Please note the similarity of Tony's slide with biological cells. Hmmm ...

* Cedric Berger: Has anyone asked AI to generate an image of a factual made of network base units?

* Michael Robbins: Agreed, Tony. And we’ve talked about this on LinkedIn. A constellation of domain-specific and ecosystem-based Community Knowledge Graphs and Language Models. #CKGs and #CLMs
** [[MikeBennett|Mike Bennett]]: TLO as mitochondria?
** Michael Robbins: Language is inseparable from culture and context

* Harvey King: Do KG's take a different nature when dealing with math?
** [[AndreaWesterinen|Andrea Westerinen]]: Not different, but with more rules?
** [[MichaelDeBellis|Michael DeBellis]]: Yes. An ontology uses explicit models like linear algebra. LLMs use linear algebra but computes answers based on examples, doesn't have a theoretical model of math (or other domains) as an ontology does.

* [[SusanneVejdemo|Sus Vejdemo]]: Thanks for a great talk! I love the conceptualization of embeddings as ontologies.

* [[MikeBennett|Mike Bennett]]: If you reify every edge you can give each one an analog value

* [[MarkUnderwood|Mark Underwood]]: For those interested in practical cybersec use cases (lots of chatty network data, some NLP), lot of narrow domain-specific emergent ontologies; e.g., Lambda / microservice mesh etc. mark.underwood@syf.com

* [[GaryBergCross|Gary Berg-Cross]]: I think the working memory graph is a useful early concept bring the LLMs and ontologies together but it is not so easy to capture what is the context for knowledge in this representation. I would guess this is a sub-set of the fluid knowledge of what human cognition employs. Much remains unconscious. But with research AI systems may make more of this explicit.

* [[RaviSharma|Ravi Sharma]]: My query is what is the relationship among reification, provenance and context history?

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Tony, “LLMs for compute” and “As much data into graph, then translating the graph paths to NL and adding to the LLM”

* [[GaryBergCross|Gary Berg-Cross]]: Reality may be atomistically discrete but at such a nano-level that continuous models make better predictions than discrete models that are orders of magnitude too gross rather than fine grained.

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Kurt: “Community Language Models - decentralized, federated, ad hoc network of information”

* [[GaryBergCross|Gary Berg-Cross]]: Models need to be more than federated. Because we center on semantic accuracy and relevance they need to be semantically harmonized.

== Resources ==
* [https://bit.ly/46DJyvo Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 18

2023-10-23T02:51:43Z

AndreaWesterinen: /* Participants */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 1]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::18 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Kurt Cagle''', Author of [https://thecaglereport.com/ The Cagle Report]
** '''Title:''' Complementary Thinking: Language Models, Ontologies and Knowledge Graphs
** '''Abstract:''' With the advent of Retrieval Augmented Generators (RAGs), a more or less standardized workflow has become available for integrating large language models such as ChatGPT with knowledge graphs. This in turn has raised the question about the nature of ontologies associated with LLMs and how knowledge graphs can be structured and queried to make integrated data access possible between the two types of systems. In this talk, Editor and AI Explorer Kurt Cagle of The Cagle Report looks at this process and discusses how they affect both knowledge portals and ontology design.
** [https://bit.ly/3S040lR Slides]
* '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/ LinkedIn])
** '''Title:''' How Ontologies Can Unlock the Potential of Large Language Models for Business
** '''Abstract:''' LLMs have remarkable capabilities; they can craft letters, analyze data, orchestrate workflows, generate code, and much more. Companies such as Google, Apple, Amazon, Meta, and Microsoft are all investing heavily in this technology. Everything indicates that LLMs have enormous disruptive potential. However, there is a problem: they can hallucinate, and for any serious business, that is a deal-breaker. This is where ontologies can come in. In combination with Knowledge Graphs, they can place guardrails around the LLMs, thus allowing organizations to harness the capabilities of LLMs within the framework of a safely controlled ontological structure.

[https://bit.ly/46A3EH2 Slides]

[https://bit.ly/46DJyvo Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 18 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=18&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* Kurt Cagle
* Tony Seale
* [[KenBaclawski|Ken Baclawski]]

== Discussion ==
* Anthony's OtterPilot: Hi, I'm an AI assistant helping Anthony Alcaraz take notes for this meeting. Follow along the transcript here: https://otter.ai/u/WLIaj2w-OmOEoVVCP5gURhhZHaY?utm_source=va_chat_link_2  You'll also be able to see screenshots of key moments, add highlights, comments, or action items to anything being said, and get an automatic summary after the meeting.
** [[MikeBennett|Mike Bennett]]: Dang! Put on a session on AI and the AIs start showing up!

* [[AndreaWesterinen|Andrea Westerinen]]: I like the quote, “Ontologies are the shapes of information and knowledge”

* [[ToddSchneider|Todd Schneider]]: What is a ‘shape of information’?
** [[AndreaWesterinen|Andrea Westerinen]]: Roles, constraints, relationships for a domain; a local representation

* [[RaviSharma|Ravi Sharma]]: Can LLMs and/or ontologies be used to detect AI artifacts?
** [[AndreaWesterinen|Andrea Westerinen]]: @Ravi Sharma There are some different articles on this, but basically no.
** [[MikeBennett|Mike Bennett]]: Not unless you could write an ontology that defines what it is to be truly human
** Michael Robbins: Or rebuild the web from the bottom up to embed new frameworks for digital identity and content provenance/authenticity (which is what we need to commit to)
** [[AndreaWesterinen|Andrea Westerinen]]: Provenance also aids with detecting bots

* [[MichaelDeBellis|Michael DeBellis]]: Someone asked about agents in relation to Data Mesh (can't find the orig comment) IMO Data Mesh could be implemented in terms of Agents but typically isn't.

* [[RaviSharma|Ravi Sharma]]: Is there one kind or multiple kinds of connectivity in KG as well as in Ontologies?

* [[RaviSharma|Ravi Sharma]]: Container in the sense of domain overlaps especially overlapping vocabs as Venn diagrams?

* Anh: Why is it hard to build LLMs for other languages? Why can't it be replicated easily when translation work (e.g. Facebook, Google Translate, LinkedIn) has already done a somewhat good job?
** [[AndreaWesterinen|Andrea Westerinen]]: It is about the training data.
** [[BartGajderowicz|Bart Gajderowicz]]: [Anh], on longer text, translation looses many of the cultural nuances of a language, and looses context. Also, most training data is in English so most models are trained on English then translated.
** Anh: Thank you, @Andrea Westerinen & @Bart Gajderowicz. Does it meean it'd cost the same to build a new LLMs for a new language?
** [[AndreaWesterinen|Andrea Westerinen]]: I would believe so.
** [[BartGajderowicz|Bart Gajderowicz]]: The volume of English text for training is cheaper, it’s just the web. So I’d imagine finding enough text in your target language would be the biggest cost. Librarians are our friends here 🙂
** [[PennyAnderson|Penny Anderson]]: this context free is difficult when you start thinking of utility of UI
** Anh: Could you please elaborate more on utility of UI? @Penny Anderson
** [[PennyAnderson|Penny Anderson]]: I mean UIs tend to be process driven data-centric is not tightly bound to a particular process that is context free
** [[PennyAnderson|Penny Anderson]]: other than search ?
** Anh: I see. Thanks.
** [[PennyAnderson|Penny Anderson]]: Data providence, Ethical AI ?
** [[AndreaWesterinen|Andrea Westerinen]]: @Penny Anderson Much better declared and reasoned against in KGs.
** [[PennyAnderson|Penny Anderson]]: Zero trust in networks?

* [[RaviSharma|Ravi Sharma]]: Kurt when you talk of box, you are essentially stating in and out of scope items or is there a way of capturing the info outside the box and bringing it in?

* [[MichaelDeBellis|Michael DeBellis]]: I don't agree that knowledge graphs are a form of data mesh. They are two related but different concepts. Data mesh is essentially an architecture for enterprise data definition and management. Knowledge graphs are a tool that can be used to implement a data mesh. Data mesh IMO brings the philosophy of microservices to data.
** Alan Morrison: Michael, re: microservices to data, should we be thinking of agents as messengers and KGs as the data resource>

* [[RaviSharma|Ravi Sharma]]: What kind of aggregates are these data shapes? Are these only valid for a class of data or can you mix data types in a shape aggregate?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi: in SHACL you define similar things as you do with OWL. E.g., does a property have to have exactly one value, the datatypes of a property, other constraints. The difference is OWL is used for reasoning over large knowledge graphs and uses the Open World Assumption. SHACL is for constraining data so it uses the Closed World Assumption. E.g., you can define ss_number as a property that must have exactly one value in either OWL or SHACL but in OWL you will almost never trigger an error if the restriction isn't satisfied due to OWA. With SHACL you will get error messages due to CWA

* [[ToddSchneider|Todd Schneider]]: LLM reasoning is probabilistic.

* Ayya Niyyanika Bhikkhuni: When weighting a concept is that some sort of credibility score?

* Michael Robbins: LangChains - I’m thinking about what role ConLangs could serve its intermediaries n revolutionizing language modes and NLP. Happy to have a follow-up discussion with anyone who is interested. https://en.wikipedia.org/wiki/Constructed_language

* [[RaviSharma|Ravi Sharma]]: What is the equivalent of Objects in LLM? What are these entities called and can same onto-entity be different in LLM context?
** [[MikeBennett|Mike Bennett]]: Words, I presume?

* [[RaviSharma|Ravi Sharma]]: Kurt thanks for including wonderful valuable background cultural images, these are inspiring. Are you also conveying the there is external (databased) and internal knowledge such as contemplation?

* [[RaviSharma|Ravi Sharma]]: Why are you limiting your examples to RDF why not MOF also?

* [[ToddSchneider|Todd Schneider]]: The important question is not mapping(s). It’s how can well constructed ontologies be used in the ingestion/training of LLMs.

* [[SusanneVejdemo|Sus Vejdemo]]: Question: what are good case studies of KGE and LLM integrations?
** [[AndreaWesterinen|Andrea Westerinen]]: I addressed some of this in the opening session. “Hybrid systems” include both where LLMs help ontologies (actually the Oct 25th session) and where ontologies help LLMs (Oct 4 and Nov 1 sessions).
** [[AndreaWesterinen|Andrea Westerinen]]: Also, Tony is highlighting a GREAT integration.

* [[MichaelDeBellis|Michael DeBellis]]: One question I have is what happens once you load a knowledge graph into an LLM? I know it can be done but once you load say a Turtle file into the LLM then what?
** [[BartGajderowicz|Bart Gajderowicz]]: Generally the LLM builds a small knowledge graph instance inside its memory and you can query it. Ask it to write SPARQL to get some instances, etc. I have not seen it used for large KGs, just small ones.

* Amit Jain: Will the recording be shared with attendees after the summit?
** [[MikeBennett|Mike Bennett]]: Yes, the recording will be uploaded to the session page when it is ready.

* Cedric Berger: Is there any metrics to measure that indeed KG combines with LLM are less hallucinating?
** [[AndreaWesterinen|Andrea Westerinen]]: I will try to provide these in the summary. I have read papers on this as well as blog posts.
** [[BartGajderowicz|Bart Gajderowicz]]: All the work I’ve come across relies on the knowledge graph to provide explicit knowledge. So if you can ground the LLM with a graph, you can verify if the answers the LLM provides are “facts” in the graph

* Kurt Cagle: Once you load in the ontology into a conversation, it will create (to some extent) an LLM conceptual space for that data. Also keep in mind that getting ALL of a knowledge graph via a RAG is usually not feasible.

* [[RaviSharma|Ravi Sharma]]: Tony great diagram for LLM ontology lop to improve each other.why is ontology weak in capturing concepts Vs LLM?

* [[MichaelDeBellis|Michael DeBellis]]: One of the most interesting papers I've read is from Lawrence Berkeley Labs on using LLMs to extend an ontology.

* [[RaviSharma|Ravi Sharma]]: How would models such as in physics work with LLM and Ontology loop or cycle that you show, actually ontologies are conceptually richer than KGs alone?
** [[AndreaWesterinen|Andrea Westerinen]]: KGs are the data, ontologies are the concepts … So, it does not seem right to ask about one being richer.
** Michael Robbins: https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its-wolfram-superpowers/

* [[RaviSharma|Ravi Sharma]]: Tony the built-in ubcertainty in LLMs gives it extra power to apply to real life probabilistic world?

* [[RaviSharma|Ravi Sharma]]: Tony LLMs and analog and ontology as Quantum? great way. thanks

* [[MikeBennett|Mike Bennett]]: Left brain v right brain is a great way of talking about ontology v LLM. Now you have to create a good corpus collosum.
** [[GaryBergCross|Gary Berg-Cross]]: A better model than left right hemispheres is by layers - old, mid brain (associative) and neo-cortex. They interconnect in many ways and some by the limbic system.
** [[BartGajderowicz|Bart Gajderowicz]]: I like the System 1 / System 2 analogy

* Cedric Berger: Aren’t LLMs also kind of discrete as relying on vectors (arrays of numbers) of limited dimensions?
** [[BartGajderowicz|Bart Gajderowicz]]: LLMs are fundamentally probabilistic, not discrete. Some models are trained to provide discrete classifications, but that’s just at the output level. Internally they are probabilistic.
** [[AndreaWesterinen|Andrea Westerinen]]: The discreteness pertains to the encoding. But, does the probabilistic nature of LLMs make it more continuous? I am not sure.
** [[MikeBennett|Mike Bennett]]: The weightings are a number rather than a logical truth value as in KGs
** Kurt Cagle: Even with KGs, you can set up reifications that also set up Bayesians that are again more fuzzy (or at least more stochastic).
** [[MikeBennett|Mike Bennett]]: If you had an ontology with weightings instead of truth values and can train those values, you have a semantic network like a brain/mind.
** Kurt Cagle: It's where I think we're heading. People in the semantic space have known for years that knowledge is fuzzy / fractal, but getting there has always been the rub.
** [[MikeBennett|Mike Bennett]]: I roughed out an idea for this kind of semantic network application back in the 90s.

* Michael Robbins: A great article on vector embeddings: https://kdb.ai/learning-hub/fundamentals/vector-embeddings/ How can we use this for transparency and explainability? Give users confidence intervals (and other potential response options) along with responses?

* [[RaviSharma|Ravi Sharma]]: What are vectors equivalent to in LLM context?

* Harvey King: Are the embeddings stored across 3 layers?
** [[AndreaWesterinen|Andrea Westerinen]]: The embeddings are there, but simplified/reduced.

* Anh: Does KG consider the time stamp of the assertions/objects? Context of my question: could we use it to mark the originality of posts of similar contents to alert plagiarism.
** [[AndreaWesterinen|Andrea Westerinen]]: The KG CAN do this, if it is encoded.

* [[RaviSharma|Ravi Sharma]]: Has anyone done this a roundtrip quality check, learn from LLM and put it in ontologies and the other way around?
** Benoit Claise: In which context/use case?
** [[MichaelDeBellis|Michael DeBellis]]: That paper I posted earlier from Lawrence Berkeley Labs used LLM to extend an ontology but just went in one direction, expanding the ontology not changing the LLM.

* [[RaviSharma|Ravi Sharma]]: Can you do reasoning on same concept in both to differentiate their respective strengths and weakness?

* [[RaviSharma|Ravi Sharma]]: Tony your tree or chain of thought are great ways

* [[RaviSharma|Ravi Sharma]]: Tony and Kurt Can you address feature space Vs training set learning approaches?

* Liju Fan: Why are the relations in the ontology explicit?
** [[AndreaWesterinen|Andrea Westerinen]]: Relations are defined, and they can be inferred, but this is the essence of ontology. Ontologies are “open world” but do need relationships.
** [[MichaelDeBellis|Michael DeBellis]]: Because in the ontology you (usually manually) create the relations. In an LLM the relations are inferred by the ML algorithm and usually can't be manually changed
** Liju Fan: It seems there is a need to be able to rename LLM inferred relations for them to be human-understandable and practically useful.

* Kurt Cagle: Please note the similarity of Tony's slide with biological cells. Hmmm ...

* Cedric Berger: Has anyone asked AI to generate an image of a factual made of network base units?

* Michael Robbins: Agreed, Tony. And we’ve talked about this on LinkedIn. A constellation of domain-specific and ecosystem-based Community Knowledge Graphs and Language Models. #CKGs and #CLMs
** [[MikeBennett|Mike Bennett]]: TLO as mitochondria?
** Michael Robbins: Language is inseparable from culture and context

* Harvey King: Do KG's take a different nature when dealing with math?
** [[AndreaWesterinen|Andrea Westerinen]]: Not different, but with more rules?
** [[MichaelDeBellis|Michael DeBellis]]: Yes. An ontology uses explicit models like linear algebra. LLMs use linear algebra but computes answers based on examples, doesn't have a theoretical model of math (or other domains) as an ontology does.

* [[SusanneVejdemo|Sus Vejdemo]]: Thanks for a great talk! I love the conceptualization of embeddings as ontologies.

* [[MikeBennett|Mike Bennett]]: If you reify every edge you can give each one an analog value

* [[MarkUnderwood|Mark Underwood]]: For those interested in practical cybersec use cases (lots of chatty network data, some NLP), lot of narrow domain-specific emergent ontologies; e.g., Lambda / microservice mesh etc. mark.underwood@syf.com

* [[GaryBergCross|Gary Berg-Cross]]: I think the working memory graph is a useful early concept bring the LLMs and ontologies together but it is not so easy to capture what is the context for knowledge in this representation. I would guess this is a sub-set of the fluid knowledge of what human cognition employs. Much remains unconscious. But with research AI systems may make more of this explicit.

* [[RaviSharma|Ravi Sharma]]: My query is what is the relationship among reification, provenance and context history?

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Tony, “LLMs for compute” and “As much data into graph, then translating the graph paths to NL and adding to the LLM”

* [[GaryBergCross|Gary Berg-Cross]]: Reality may be atomistically discrete but at such a nano-level that continuous models make better predictions than discrete models that are orders of magnitude too gross rather than fine grained.

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Kurt: “Community Language Models - decentralized, federated, ad hoc network of information”

* [[GaryBergCross|Gary Berg-Cross]]: Models need to be more than federated. Because we center on semantic accuracy and relevance they need to be semantically harmonized.

== Resources ==
* [https://bit.ly/46DJyvo Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 18

2023-10-23T02:51:06Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::A look across the industry, Part 1]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::18 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Convener
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==
* '''Kurt Cagle''', Author of [https://thecaglereport.com/ The Cagle Report]
** '''Title:''' Complementary Thinking: Language Models, Ontologies and Knowledge Graphs
** '''Abstract:''' With the advent of Retrieval Augmented Generators (RAGs), a more or less standardized workflow has become available for integrating large language models such as ChatGPT with knowledge graphs. This in turn has raised the question about the nature of ontologies associated with LLMs and how knowledge graphs can be structured and queried to make integrated data access possible between the two types of systems. In this talk, Editor and AI Explorer Kurt Cagle of The Cagle Report looks at this process and discusses how they affect both knowledge portals and ontology design.
** [https://bit.ly/3S040lR Slides]
* '''Tony Seale''', Knowledge graph architect and thought leader ([https://www.linkedin.com/in/tonyseale/ LinkedIn])
** '''Title:''' How Ontologies Can Unlock the Potential of Large Language Models for Business
** '''Abstract:''' LLMs have remarkable capabilities; they can craft letters, analyze data, orchestrate workflows, generate code, and much more. Companies such as Google, Apple, Amazon, Meta, and Microsoft are all investing heavily in this technology. Everything indicates that LLMs have enormous disruptive potential. However, there is a problem: they can hallucinate, and for any serious business, that is a deal-breaker. This is where ontologies can come in. In combination with Knowledge Graphs, they can place guardrails around the LLMs, thus allowing organizations to harness the capabilities of LLMs within the framework of a safely controlled ontological structure.

[https://bit.ly/46A3EH2 Slides]

[https://bit.ly/46DJyvo Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 18 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=18&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==

== Discussion ==
* Anthony's OtterPilot: Hi, I'm an AI assistant helping Anthony Alcaraz take notes for this meeting. Follow along the transcript here: https://otter.ai/u/WLIaj2w-OmOEoVVCP5gURhhZHaY?utm_source=va_chat_link_2  You'll also be able to see screenshots of key moments, add highlights, comments, or action items to anything being said, and get an automatic summary after the meeting.
** [[MikeBennett|Mike Bennett]]: Dang! Put on a session on AI and the AIs start showing up!

* [[AndreaWesterinen|Andrea Westerinen]]: I like the quote, “Ontologies are the shapes of information and knowledge”

* [[ToddSchneider|Todd Schneider]]: What is a ‘shape of information’?
** [[AndreaWesterinen|Andrea Westerinen]]: Roles, constraints, relationships for a domain; a local representation

* [[RaviSharma|Ravi Sharma]]: Can LLMs and/or ontologies be used to detect AI artifacts?
** [[AndreaWesterinen|Andrea Westerinen]]: @Ravi Sharma There are some different articles on this, but basically no.
** [[MikeBennett|Mike Bennett]]: Not unless you could write an ontology that defines what it is to be truly human
** Michael Robbins: Or rebuild the web from the bottom up to embed new frameworks for digital identity and content provenance/authenticity (which is what we need to commit to)
** [[AndreaWesterinen|Andrea Westerinen]]: Provenance also aids with detecting bots

* [[MichaelDeBellis|Michael DeBellis]]: Someone asked about agents in relation to Data Mesh (can't find the orig comment) IMO Data Mesh could be implemented in terms of Agents but typically isn't.

* [[RaviSharma|Ravi Sharma]]: Is there one kind or multiple kinds of connectivity in KG as well as in Ontologies?

* [[RaviSharma|Ravi Sharma]]: Container in the sense of domain overlaps especially overlapping vocabs as Venn diagrams?

* Anh: Why is it hard to build LLMs for other languages? Why can't it be replicated easily when translation work (e.g. Facebook, Google Translate, LinkedIn) has already done a somewhat good job?
** [[AndreaWesterinen|Andrea Westerinen]]: It is about the training data.
** [[BartGajderowicz|Bart Gajderowicz]]: [Anh], on longer text, translation looses many of the cultural nuances of a language, and looses context. Also, most training data is in English so most models are trained on English then translated.
** Anh: Thank you, @Andrea Westerinen & @Bart Gajderowicz. Does it meean it'd cost the same to build a new LLMs for a new language?
** [[AndreaWesterinen|Andrea Westerinen]]: I would believe so.
** [[BartGajderowicz|Bart Gajderowicz]]: The volume of English text for training is cheaper, it’s just the web. So I’d imagine finding enough text in your target language would be the biggest cost. Librarians are our friends here 🙂
** [[PennyAnderson|Penny Anderson]]: this context free is difficult when you start thinking of utility of UI
** Anh: Could you please elaborate more on utility of UI? @Penny Anderson
** [[PennyAnderson|Penny Anderson]]: I mean UIs tend to be process driven data-centric is not tightly bound to a particular process that is context free
** [[PennyAnderson|Penny Anderson]]: other than search ?
** Anh: I see. Thanks.
** [[PennyAnderson|Penny Anderson]]: Data providence, Ethical AI ?
** [[AndreaWesterinen|Andrea Westerinen]]: @Penny Anderson Much better declared and reasoned against in KGs.
** [[PennyAnderson|Penny Anderson]]: Zero trust in networks?

* [[RaviSharma|Ravi Sharma]]: Kurt when you talk of box, you are essentially stating in and out of scope items or is there a way of capturing the info outside the box and bringing it in?

* [[MichaelDeBellis|Michael DeBellis]]: I don't agree that knowledge graphs are a form of data mesh. They are two related but different concepts. Data mesh is essentially an architecture for enterprise data definition and management. Knowledge graphs are a tool that can be used to implement a data mesh. Data mesh IMO brings the philosophy of microservices to data.
** Alan Morrison: Michael, re: microservices to data, should we be thinking of agents as messengers and KGs as the data resource>

* [[RaviSharma|Ravi Sharma]]: What kind of aggregates are these data shapes? Are these only valid for a class of data or can you mix data types in a shape aggregate?
** [[MichaelDeBellis|Michael DeBellis]]: Ravi: in SHACL you define similar things as you do with OWL. E.g., does a property have to have exactly one value, the datatypes of a property, other constraints. The difference is OWL is used for reasoning over large knowledge graphs and uses the Open World Assumption. SHACL is for constraining data so it uses the Closed World Assumption. E.g., you can define ss_number as a property that must have exactly one value in either OWL or SHACL but in OWL you will almost never trigger an error if the restriction isn't satisfied due to OWA. With SHACL you will get error messages due to CWA

* [[ToddSchneider|Todd Schneider]]: LLM reasoning is probabilistic.

* Ayya Niyyanika Bhikkhuni: When weighting a concept is that some sort of credibility score?

* Michael Robbins: LangChains - I’m thinking about what role ConLangs could serve its intermediaries n revolutionizing language modes and NLP. Happy to have a follow-up discussion with anyone who is interested. https://en.wikipedia.org/wiki/Constructed_language

* [[RaviSharma|Ravi Sharma]]: What is the equivalent of Objects in LLM? What are these entities called and can same onto-entity be different in LLM context?
** [[MikeBennett|Mike Bennett]]: Words, I presume?

* [[RaviSharma|Ravi Sharma]]: Kurt thanks for including wonderful valuable background cultural images, these are inspiring. Are you also conveying the there is external (databased) and internal knowledge such as contemplation?

* [[RaviSharma|Ravi Sharma]]: Why are you limiting your examples to RDF why not MOF also?

* [[ToddSchneider|Todd Schneider]]: The important question is not mapping(s). It’s how can well constructed ontologies be used in the ingestion/training of LLMs.

* [[SusanneVejdemo|Sus Vejdemo]]: Question: what are good case studies of KGE and LLM integrations?
** [[AndreaWesterinen|Andrea Westerinen]]: I addressed some of this in the opening session. “Hybrid systems” include both where LLMs help ontologies (actually the Oct 25th session) and where ontologies help LLMs (Oct 4 and Nov 1 sessions).
** [[AndreaWesterinen|Andrea Westerinen]]: Also, Tony is highlighting a GREAT integration.

* [[MichaelDeBellis|Michael DeBellis]]: One question I have is what happens once you load a knowledge graph into an LLM? I know it can be done but once you load say a Turtle file into the LLM then what?
** [[BartGajderowicz|Bart Gajderowicz]]: Generally the LLM builds a small knowledge graph instance inside its memory and you can query it. Ask it to write SPARQL to get some instances, etc. I have not seen it used for large KGs, just small ones.

* Amit Jain: Will the recording be shared with attendees after the summit?
** [[MikeBennett|Mike Bennett]]: Yes, the recording will be uploaded to the session page when it is ready.

* Cedric Berger: Is there any metrics to measure that indeed KG combines with LLM are less hallucinating?
** [[AndreaWesterinen|Andrea Westerinen]]: I will try to provide these in the summary. I have read papers on this as well as blog posts.
** [[BartGajderowicz|Bart Gajderowicz]]: All the work I’ve come across relies on the knowledge graph to provide explicit knowledge. So if you can ground the LLM with a graph, you can verify if the answers the LLM provides are “facts” in the graph

* Kurt Cagle: Once you load in the ontology into a conversation, it will create (to some extent) an LLM conceptual space for that data. Also keep in mind that getting ALL of a knowledge graph via a RAG is usually not feasible.

* [[RaviSharma|Ravi Sharma]]: Tony great diagram for LLM ontology lop to improve each other.why is ontology weak in capturing concepts Vs LLM?

* [[MichaelDeBellis|Michael DeBellis]]: One of the most interesting papers I've read is from Lawrence Berkeley Labs on using LLMs to extend an ontology.

* [[RaviSharma|Ravi Sharma]]: How would models such as in physics work with LLM and Ontology loop or cycle that you show, actually ontologies are conceptually richer than KGs alone?
** [[AndreaWesterinen|Andrea Westerinen]]: KGs are the data, ontologies are the concepts … So, it does not seem right to ask about one being richer.
** Michael Robbins: https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its-wolfram-superpowers/

* [[RaviSharma|Ravi Sharma]]: Tony the built-in ubcertainty in LLMs gives it extra power to apply to real life probabilistic world?

* [[RaviSharma|Ravi Sharma]]: Tony LLMs and analog and ontology as Quantum? great way. thanks

* [[MikeBennett|Mike Bennett]]: Left brain v right brain is a great way of talking about ontology v LLM. Now you have to create a good corpus collosum.
** [[GaryBergCross|Gary Berg-Cross]]: A better model than left right hemispheres is by layers - old, mid brain (associative) and neo-cortex. They interconnect in many ways and some by the limbic system.
** [[BartGajderowicz|Bart Gajderowicz]]: I like the System 1 / System 2 analogy

* Cedric Berger: Aren’t LLMs also kind of discrete as relying on vectors (arrays of numbers) of limited dimensions?
** [[BartGajderowicz|Bart Gajderowicz]]: LLMs are fundamentally probabilistic, not discrete. Some models are trained to provide discrete classifications, but that’s just at the output level. Internally they are probabilistic.
** [[AndreaWesterinen|Andrea Westerinen]]: The discreteness pertains to the encoding. But, does the probabilistic nature of LLMs make it more continuous? I am not sure.
** [[MikeBennett|Mike Bennett]]: The weightings are a number rather than a logical truth value as in KGs
** Kurt Cagle: Even with KGs, you can set up reifications that also set up Bayesians that are again more fuzzy (or at least more stochastic).
** [[MikeBennett|Mike Bennett]]: If you had an ontology with weightings instead of truth values and can train those values, you have a semantic network like a brain/mind.
** Kurt Cagle: It's where I think we're heading. People in the semantic space have known for years that knowledge is fuzzy / fractal, but getting there has always been the rub.
** [[MikeBennett|Mike Bennett]]: I roughed out an idea for this kind of semantic network application back in the 90s.

* Michael Robbins: A great article on vector embeddings: https://kdb.ai/learning-hub/fundamentals/vector-embeddings/ How can we use this for transparency and explainability? Give users confidence intervals (and other potential response options) along with responses?

* [[RaviSharma|Ravi Sharma]]: What are vectors equivalent to in LLM context?

* Harvey King: Are the embeddings stored across 3 layers?
** [[AndreaWesterinen|Andrea Westerinen]]: The embeddings are there, but simplified/reduced.

* Anh: Does KG consider the time stamp of the assertions/objects? Context of my question: could we use it to mark the originality of posts of similar contents to alert plagiarism.
** [[AndreaWesterinen|Andrea Westerinen]]: The KG CAN do this, if it is encoded.

* [[RaviSharma|Ravi Sharma]]: Has anyone done this a roundtrip quality check, learn from LLM and put it in ontologies and the other way around?
** Benoit Claise: In which context/use case?
** [[MichaelDeBellis|Michael DeBellis]]: That paper I posted earlier from Lawrence Berkeley Labs used LLM to extend an ontology but just went in one direction, expanding the ontology not changing the LLM.

* [[RaviSharma|Ravi Sharma]]: Can you do reasoning on same concept in both to differentiate their respective strengths and weakness?

* [[RaviSharma|Ravi Sharma]]: Tony your tree or chain of thought are great ways

* [[RaviSharma|Ravi Sharma]]: Tony and Kurt Can you address feature space Vs training set learning approaches?

* Liju Fan: Why are the relations in the ontology explicit?
** [[AndreaWesterinen|Andrea Westerinen]]: Relations are defined, and they can be inferred, but this is the essence of ontology. Ontologies are “open world” but do need relationships.
** [[MichaelDeBellis|Michael DeBellis]]: Because in the ontology you (usually manually) create the relations. In an LLM the relations are inferred by the ML algorithm and usually can't be manually changed
** Liju Fan: It seems there is a need to be able to rename LLM inferred relations for them to be human-understandable and practically useful.

* Kurt Cagle: Please note the similarity of Tony's slide with biological cells. Hmmm ...

* Cedric Berger: Has anyone asked AI to generate an image of a factual made of network base units?

* Michael Robbins: Agreed, Tony. And we’ve talked about this on LinkedIn. A constellation of domain-specific and ecosystem-based Community Knowledge Graphs and Language Models. #CKGs and #CLMs
** [[MikeBennett|Mike Bennett]]: TLO as mitochondria?
** Michael Robbins: Language is inseparable from culture and context

* Harvey King: Do KG's take a different nature when dealing with math?
** [[AndreaWesterinen|Andrea Westerinen]]: Not different, but with more rules?
** [[MichaelDeBellis|Michael DeBellis]]: Yes. An ontology uses explicit models like linear algebra. LLMs use linear algebra but computes answers based on examples, doesn't have a theoretical model of math (or other domains) as an ontology does.

* [[SusanneVejdemo|Sus Vejdemo]]: Thanks for a great talk! I love the conceptualization of embeddings as ontologies.

* [[MikeBennett|Mike Bennett]]: If you reify every edge you can give each one an analog value

* [[MarkUnderwood|Mark Underwood]]: For those interested in practical cybersec use cases (lots of chatty network data, some NLP), lot of narrow domain-specific emergent ontologies; e.g., Lambda / microservice mesh etc. mark.underwood@syf.com

* [[GaryBergCross|Gary Berg-Cross]]: I think the working memory graph is a useful early concept bring the LLMs and ontologies together but it is not so easy to capture what is the context for knowledge in this representation. I would guess this is a sub-set of the fluid knowledge of what human cognition employs. Much remains unconscious. But with research AI systems may make more of this explicit.

* [[RaviSharma|Ravi Sharma]]: My query is what is the relationship among reification, provenance and context history?

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Tony, “LLMs for compute” and “As much data into graph, then translating the graph paths to NL and adding to the LLM”

* [[GaryBergCross|Gary Berg-Cross]]: Reality may be atomistically discrete but at such a nano-level that continuous models make better predictions than discrete models that are orders of magnitude too gross rather than fine grained.

* [[AndreaWesterinen|Andrea Westerinen]]: Quote from Kurt: “Community Language Models - decentralized, federated, ad hoc network of information”

* [[GaryBergCross|Gary Berg-Cross]]: Models need to be more than federated. Because we center on semantic accuracy and relevance they need to be semantically harmonized.

== Resources ==
* [https://bit.ly/46DJyvo Video Recording]

== Previous Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[<<ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=desc|limit=3}}

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_18]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 04

2023-10-23T02:03:45Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Overview]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::4 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Conveners
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==

'''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''

'''Title:''' ''Fall Series Kickoff and Overview''

'''Abstract:''' The opening session of the Ontology Summit 2024 Fall Series overviews the LLM, ontology and knowledge graph landscapes, as well as introducing the participating speakers. The goal of the Series is to understand, discuss and debate the similarities, differences and overlaps across these landscapes. In addition, we will use these sessions to help to formulate the full 2024 Summit.

[https://bit.ly/3Q28U00 Slides]

[https://bit.ly/3rCnyC0 Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 4 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=04&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[AlexShkotin|Alex Shkotin]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[ToddSchneider|Todd Schneider]]
* [[MikeBennett|Mike Bennett]]
* Ayya Niyyanika Bhikkhuni
* Bill McCarthy
* Zefi Kavvadia
* [[RamSriram|Ram D Sriram]]
* Andrew McCaffrey
* Steve Wartik
* [[BartGajderowicz|Bart Gajderowicz]]
* [[MarkFox|Mark Fox]]
* Seungmin Seo
* JL Valente
* [[MichaelDeBellis|Michael DeBellis]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* Sima Yazdani
* [[JohnSowa|John Sowa]]
* [[KenBaclawski|Ken Baclawski]]
* [[RaviSharma|Ravi Sharma]]
* Sergey Rodionov
* Taj Uddin
* [[MarkRessler|Mark Ressler]]
* Asiyah Yu Lin
* Hayden Spence
* Michael Singer
* Roberta Ferrario
* Chris Novell
* Emanuele Bottazzi
* Marco Monti
* [[JanetSinger|Janet Singer]]

== Discussion ==
* [[MikeBennett|Mike Bennett]]: Andrea's quote: "Ontologies are the backing definitions behind knowledge graphs" is a great way of describing the distinction between them.
** Emanuele Bottazzi: Or justifications

* [[ToddSchneider|Todd Schneider]]: Many Knowledge Graphs are not based on an ontology.
** [[AlexShkotin|Alex Shkotin]]: but keep it inside
** [[MikeBennett|Mike Bennett]]: I would not characterize such a thing as a knowledge graph, even if it re-uses that label for itself. Whence the claim of 'Knowledge' in KG if not semantics? Might not be an OWL-ology of course.
** [[BartGajderowicz|Bart Gajderowicz]]: An ontology is the “schema” for a knowledge graph, so it may not be designed well but there is a “schema” that defines nodes and edges in some way.

* Steven Wartik: I like to distinguish between a KG and a knowledge base. A KG is a graph. It doesn't necessarily have a schema. A KB is a KG whose schema is an ontology. This is just terminology, but I find it helps my sponsors understand.

* [[AlexShkotin|Alex Shkotin]]: Give me KG and I extract it's ontology.

* [[KenBaclawski|Ken Baclawski]]: KGs were covered in Ontology Summit 2020. The communique has precise definitions: https://ontologforum.s3.amazonaws.com/OntologySummit2020/Communique/OntologySummit2020Communique.pdf

* [[ToddSchneider|Todd Schneider]]: Knowledge =def. “facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject” (from New Oxford American Dictionary)

* [[ToddSchneider|Todd Schneider]]: ‘Meaning’ is an ambiguous term.

* Andrew McCaffrey: To "table" a motion means completely the opposite things in the US and the UK. :D

* Ayya Niyyanika Bhikkhuni: As generative AI hallucinations become an issue, there seems a need for credibility scoring. I am about a decade out-of-the-loop, but know we were talking about this many summits ago. This is in regards to trust.
** [[BartGajderowicz|Bart Gajderowicz]]: Explanations WITH hallucinations are a huge problem for LLMs. They sound credible, and may be logically sound, but are completely wrong.
** Emanuele Bottazzi: Perhaps all the probabilistic approaches cannot be explanatory, since they “happen” to be wrong or right
** [[BartGajderowicz|Bart Gajderowicz]]: Ideally the explanation would come from explicit knowledge. Most LLMs just don’t have that. Ensemble ML architectures may include explicit knowledge somewhere, but if the underlying processes and representations are probabilistic we reach a hard limit on explainability. Of course you can have an explanation that provides “certainty” about the answer and explanation, which is often sufficient.
** Emanuele Bottazzi: I would add that ideally the explanation would come from the explicit _use_ of knowledge and principles

* [[BartGajderowicz|Bart Gajderowicz]]: Do LLMs perform natural language understanding (NLU), or just processing (NLP)?
** [[BartGajderowicz|Bart Gajderowicz]]: Given my definition of knowledge I’d say NLP only. Even a simple Word2vec embedding is able to identify similarity between complex objects, but I would not consider it understanding (or knowledge)

* Ayya Niyyanika Bhikkhuni: “What is really true” is the underlying question when translating ancient text. The project I am working on is taking translations from humans and Generative AI and it is hoped then that people practicing according to their interpretation of the texts would tune the translations based on ‘tacit knowledge.’

* Marco Monti: QUESTION: if neither LLM models nor Knowledge Graphs allow for compositionality and high contextualization of answers from a chat bot, what are the mechanisms behind the scenes of GPT X to answer so punctually and contextually ?

* [[JanetSinger|Janet Singer]]: Yes, mimicry is the key characterization of what LLMs do. Parallels the 1950s it was thought that mimicry of biological behavior would inevitably lead to a structural model of living systems, and then to artificially generated life itself. See critiques by Robert Rosen.

* [[GaryBergCross|Gary Berg-Cross]]: LLM based systems can learn on the job although you wouldn't call it based on experience. This has been said about the learning: "When a user interacts with an LLM-based system, the system is able to observe the user's responses and learn from them. This allows the system to improve its ability to generate responses that are relevant to the user's needs.

* [[GaryBergCross|Gary Berg-Cross]]: There are a number of ways that LLM-based systems can be trained using chat responses. One common approach is to use reinforcement learning. In reinforcement learning, the system is rewarded for generating responses that are positive and helpful. This encourages the system to learn what kinds of responses are most likely to be well-received by users."

* [[ToddSchneider|Todd Schneider]]: Could explain “links in OWL are not first class objects”?
** Steven Wartik: Todd, a first-class object is uniquely identifiable. A reified triple is a 1st-class object.
** Asiyah Yu Lin: I think the knowledge graph users who doesn't care too much about OWL thinking of data level or instance level. The ontology is really about classes. There is a blurred line between what is data and what is class.
** [[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Suppose you have a model of a highway as a graph where nodes are cities and links are roads. You want to model the time it takes to get from two nodes as information directly on the link. You can do that with Neo4J but now with OWL. With OWL you need to use the design pattern where you reify the relation with a new class.
** [[ToddSchneider|Todd Schneider]]: Michael, thank you for the explanation. Per your example, it could be the case that the representation (of the entities and their relations) was inadequate to support the query (i.e. with reification). Typo “ with reification’ should be ‘Without reification).
** [[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Yes. My question is how easy is it to take an OWL ontology where you have reified the relations and use graph theoretic algorithms? I don't know because I haven't used these algorithms in a long time. One thing I'm thinking about is creating an extension to OWL (I mean things like new classes and Python or SPARQL) where when you assert a new property value you have the option to create an instance of a Relation class and store data directly on that instance. That way you could treat the OWL ontology as a true graph.
** [[MichaelDeBellis|Michael DeBellis]]: Often you can even ask GPT-4 to create the KIF or CycL or CLIF .. and it will

* Hayden Spence: RE: Generating ontologies with LLMs: https://github.com/monarch-initiative/ontogpt

* [[JanetSinger|Janet Singer]]: Here, mimicry of knowledge-driven behavior is being promoted as inevitably leading to structural models of knowledge and then to ‘emergent consciousness’. Ontologies are structural (good for modeling within their scope); LLMs are behavioral
** [[JanetSinger|Janet Singer]]: Here as in the hype cycle, not by Andrea 🙂

* [[Douglas Miles|Douglas Miles]]: GPT-3 btw seems useless compared to GPT-4 on this front

* Hayden Spence: From my understanding, GPT-4 is multimodal and multimodel in the sense its training is higher parameter, it incorporates more than just text data, and the actual interface is the interaction of multiple GPT models working together.

*[[ToddSchneider|Todd Schneider]]: What is ‘semantic understanding’?

* Hayden Spence: Is the use of established controlled vocabularies that are under license like SNOMED CT, MedDRA, ICD10/0, or standards like FHIR, and the mappings between them -- once embedded -- still restricted? At what point does transformation of information collection become its own separate from the digested information.

* [[Douglas Miles|Douglas Miles]]: i don't have a question at this point.. but love this talk!

* [[JanetSinger|Janet Singer]]: Symbolic and connectionist theories of cognition are both computationalist. Leaves out 4-E embodied cognition perspective

== Resources ==
* [https://bit.ly/3rCnyC0 Video Recording]

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_04]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 25

2023-10-23T01:14:26Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 10 25

2023-10-22T18:03:39Z

AndreaWesterinen: /* Agenda */

ConferenceCall 2023 10 18

2023-10-22T17:25:39Z

AndreaWesterinen: /* Ontology Summit 2024 {{#show:{{PAGENAME}}|?session}} */

ConferenceCall 2023 10 04

2023-10-17T16:36:43Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Overview]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::4 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Conveners
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==

'''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''

'''Title:''' ''Fall Series Kickoff and Overview''

'''Abstract:''' The opening session of the Ontology Summit 2024 Fall Series overviews the LLM, ontology and knowledge graph landscapes, as well as introducing the participating speakers. The goal of the Series is to understand, discuss and debate the similarities, differences and overlaps across these landscapes. In addition, we will use these sessions to help to formulate the full 2024 Summit.

[https://bit.ly/3Q28U00 Slides]

[https://bit.ly/3rCnyC0 Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 4 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=04&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[AlexShkotin|Alex Shkotin]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[ToddSchneider|Todd Schneider]]
* [[MikeBennett|Mike Bennett]]
* Ayya Niyyanika Bhikkhuni
* Bill McCarthy
* Zefi Kavvadia
* [[RamSriram|Ram D Sriram]]
* Andrew McCaffrey
* Steve Wartik
* [[BartGajderowicz|Bart Gajderowicz]]
* [[MarkFox|Mark Fox]]
* Seungmin Seo
* JL Valente
* [[MichaelDeBellis|Michael DeBellis]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* Sima Yazdani
* [[JohnSowa|John Sowa]]
* [[KenBaclawski|Ken Baclawski]]
* [[RaviSharma|Ravi Sharma]]
* Sergey Rodionov
* Taj Uddin
* [[MarkRessler|Mark Ressler]]
* Asiyah Yu Lin
* Hayden Spence
* Michael Singer
* Roberta Ferrario
* Chris Novell
* Emanuele Bottazzi
* Marco Monti
* [[JanetSinger|Janet Singer]]

== Discussion ==
[[MikeBennett|Mike Bennett]]: Andrea's quote: "Ontologies are the backing definitions behind knowledge graphs" is a
great way of describing the distinction between them.

Emanuele Bottazzi: Or justifications

[[ToddSchneider|Todd Schneider]]: Many Knowledge Graphs are not based on an ontology.

[[AlexShkotin|Alex Shkotin]]: but keep it inside

[[MikeBennett|Mike Bennett]]: I would not characterize such a thing as a knowledge graph, even if it re-uses that label for itself. Whence the claim of 'Knowledge' in KG if not semantics? Might not be an OWL-ology of course.

[[BartGajderowicz|Bart Gajderowicz]]: An ontology is the “schema” for a knowledge graph, so it may not be designed well but there is a “schema” that defines nodes and edges in some way.

Steven Wartik: I like to distinguish between a KG and a knowledge base. A KG is a graph. It doesn't necessarily have a schema. A KB is a KG whose schema is an ontology. This is just terminology, but I find it helps my sponsors understand.

[[AlexShkotin|Alex Shkotin]]: Give me KG and I extract it's ontology.

[[KenBaclawski|Ken Baclawski]]: KGs were covered in Ontology Summit 2020. The communique has precise definitions: https://ontologforum.s3.amazonaws.com/OntologySummit2020/Communique/OntologySummit2020Communique.pdf

[[ToddSchneider|Todd Schneider]]: Knowledge =def. “facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject” (from New Oxford American Dictionary)

[[ToddSchneider|Todd Schneider]]: ‘Meaning’ is an ambiguous term.

Andrew McCaffrey: To "table" a motion means completely the opposite things in the US and the UK. :D

Ayya Niyyanika Bhikkhuni: As generative AI hallucinations become an issue, there seems a need for credibility scoring. I am about a decade out-of-the-loop, but know we were talking about this many summits ago. This is in regards to trust.

[[BartGajderowicz|Bart Gajderowicz]]: Explanations WITH hallucinations are a huge problem for LLMs. They sound credible, and may be logically sound, but are completely wrong.

Emanuele Bottazzi: Perhaps all the probabilistic approaches cannot be explanatory, since the “happen” to be wrong or right

[[BartGajderowicz|Bart Gajderowicz]]: Ideally the explanation would come from explicit knowledge. Most LLMs just don’t have that. Ensemble ML architectures may include explicit knowledge somewhere, but if the underlying processes and representations are probabilistic we reach a hard limit on explainability. Of course you can have an explanation that provides “certainty” about the answer and explanation, which is often sufficient.

Emanuele Bottazzi: I would add that ideally the explanation would come from the explicit _use_ of knowledge and principles

[[BartGajderowicz|Bart Gajderowicz]]: Do LLMs perform natural language understanding (NLU), or just processing (NLP)?

[[BartGajderowicz|Bart Gajderowicz]]: Given my definition of knowledge I’d say NLP only. Even a simple Word2vec embedding is able to identify similarity between complex objects, but I would not consider it understanding (or knowledge)

Ayya Niyyanika Bhikkhuni: “What is really true” is the underlying question when translating ancient text. The project I am working on is taking translations from humans and Generative AI and it is hoped then that people practicing according to their interpretation of the texts would tune the translations based on ‘tacit knowledge.’

Marco Monti: QUESTION: if neither LLM models nor Knowledge Graphs allow for compositionality and high contextualization of answers from a chat bot, what are the mechanisms behind the scenes of GPT X to answer so punctually and contextually ?

[[JanetSinger|Janet Singer]]: Yes, mimicry is the key characterization of what LLMs do. Parallels the 1950s it was thought that mimicry of biological behavior would inevitably lead to a structural model of living systems, and then to artificially generated life itself. See critiques by Robert Rosen.

[[GaryBergCross|Gary Berg-Cross]]: LLM based systems can learn on the job although you wouldn't call it based on experience. This has been said about the learning: "When a user interacts with an LLM-based system, the system is able to observe the user's responses and learn from them. This allows the system to improve its ability to generate responses that are relevant to the user's needs.

[[GaryBergCross|Gary Berg-Cross]]: There are a number of ways that LLM-based systems can be trained using chat responses. One common approach is to use reinforcement learning. In reinforcement learning, the system is rewarded for generating responses that are positive and helpful. This encourages the system to learn what kinds of responses are most likely to be well-received by users."

[[ToddSchneider|Todd Schneider]]: Could explain “links in OWL are not first class objects”?

Steven Wartik: Todd, a first-class object is uniquely identifiable. A reified triple is a 1st-class object.

Asiyah Yu Lin: I think the knowledge graph users who doesn't care too much about OWL thinking of data level or instance level. The ontology is really about classes. There is a blurred line between what is data and what is class.

[[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Suppose you have a model of a highway as a graph where nodes are cities and links are roads. You want to model the time it takes to get from two nodes as information directly on the link. You can do that with Neo4J but now with OWL. With OWL you need to use the design pattern where you reify the relation with a new class.

[[ToddSchneider|Todd Schneider]]: Michael, thank you for the explanation. Per your example, it could be the case that the representation (of the entities and their relations) was inadequate to support the query (i.e. with reification).
Typo “ with reification’ should be ‘Without reification).

[[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Yes. My question is how easy is it to take an OWL ontology where you have reified the relations and use graph theoretic algorithms? I don't know because I haven't used these algorithms in a long time. One thing I'm thinking about is creating an extension to OWL (I mean things like new classes and Python or SPARQL) where when you assert a new property value you have the option to create an instance of a Relation class and store data directly on that instance. That way you could treat the OWL ontology as a true graph.

[[MichaelDeBellis|Michael DeBellis]]: Often you can even ask GPT-4 to create the KIF or CycL or CLIF .. and it will

Hayden Spence: RE: Generating ontologies with LLMs: https://github.com/monarch-initiative/ontogpt

[[JanetSinger|Janet Singer]]: Here, mimicry of knowledge-driven behavior is being promoted as inevitably leading to structural models of knowledge and then to ‘emergent consciousness’. Ontologies are structural (good for modeling within their
scope); LLMs are behavioral

[[JanetSinger|Janet Singer]]: Here as in the hype cycle, not by Andrea 🙂

[[Douglas Miles|Douglas Miles]]: GPT-3 btw seems useless compared to GPT-4 on this front

Hayden Spence: From my understanding, GPT-4 is multimodal and multimodel in the sense its training is higher parameter, it incorporates more than just text data, and the actual interface is the interaction of multiple GPT models working together.

[[ToddSchneider|Todd Schneider]]: What is ‘semantic understanding’?

Hayden Spence: Is the use of established controlled vocabularies that are under license like SNOMED CT, MedDRA, ICD10/0, or standards like FHIR, and the mappings between them -- once embedded -- still restricted? At what point does transformation of information collection become its own separate from the digested information.

[[Douglas Miles|Douglas Miles]]: i don't have a question at this point.. but love this talk!

[[BartGajderowicz|Bart Gajderowicz]]: Thank you every, looking forward to the next sessions

[[JanetSinger|Janet Singer]]: Symbolic and connectionist theories of cognition are both computationalist. Leaves out 4-E embodied cognition perspective

== Resources ==
* [https://bit.ly/3rCnyC0 Video Recording]

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_04]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 04

2023-10-15T21:47:56Z

AndreaWesterinen: /* Resources */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Overview]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::4 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Conveners
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==

'''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''

'''Title:''' ''Fall Series Kickoff and Overview''

'''Abstract:''' The opening session of the Ontology Summit 2024 Fall Series overviews the LLM, ontology and knowledge graph landscapes, as well as introducing the participating speakers. The goal of the Series is to understand, discuss and debate the similarities, differences and overlaps across these landscapes. In addition, we will use these sessions to help to formulate the full 2024 Summit.

[https://bit.ly/3Q28U00 Slides]

[https://bit.ly/3rCnyC0 Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 4 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=04&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[AlexShkotin|Alex Shkotin]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[ToddSchneider|Todd Schneider]]
* [[MikeBennett|Mike Bennett]]
* Ayya Niyyanika Bhikkhuni
* Bill McCarthy
* Zefi Kavvadia
* [[RamSriram|Ram D Sriram]]
* Andrew McCaffrey
* Steve Wartik
* [[BartGajderowicz|Bart Gajderowicz]]
* [[MarkFox|Mark Fox]]
* Seungmin Seo
* JL Valente
* [[MichaelDeBellis|Michael DeBellis]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* Sima Yazdani
* [[JohnSowa|John Sowa]]
* [[KenBaclawski|Ken Baclawski]]
* [[RaviSharma|Ravi Sharma]]
* Sergey Rodionov
* Taj Uddin
* [[MarkRessler|Mark Ressler]]
* Asiyah Yu Lin
* Hayden Spence
* Michael Singer
* Roberta Ferrario
* Chris Novell
* Emanuele Bottazzi
* Marco Monti
* [[JanetSinger|Janet Singer]]

== Discussion ==
[[MikeBennett|Mike Bennett]]: I would not characterize a graph as a knowledge graph, even if it re-uses that label for itself. Whence the claim of 'Knowledge' in KG if not semantics? Might not be an OWL-ology of course.

[[BartGajderowicz|Bart Gajderowicz]]: An ontology is the “schema” for a knowledge graph, so it may not be designed well but there is a “schema” that defines nodes and edges in some way.

Steven Wartik: I like to distinguish between a KG and a knowledge base. A KG is a graph. It doesn't necessarily have a schema. A KB is a KG whose schema is an ontology. This is just terminology, but I find it helps my sponsors understand.

[[AlexShkotin|Alex Shkotin]]: Give me KG and I extract it's ontology.

[[KenBaclawski|Ken Baclawski]]: KGs were covered in Ontology Summit 2020. The communique has precise definitions: https://ontologforum.s3.amazonaws.com/OntologySummit2020/Communique/OntologySummit2020Communique.pdf

[[ToddSchneider|Todd Schneider]]: Knowledge =def. “facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject” (from New Oxford American Dictionary)

[[ToddSchneider|Todd Schneider]]: ‘Meaning’ is an ambiguous term.

Andrew McCaffrey: To "table" a motion means completely the opposite things in the US and the UK. :D

Ayya Niyyanika Bhikkhuni: As generative AI hallucinations become an issue, there seems a need for credibility scoring. I am about a decade out-of-the-loop, but know we were talking about this many summits ago. This is in regards to trust.

[[BartGajderowicz|Bart Gajderowicz]]: Explanations WITH hallucinations are a huge problem for LLMs. They sound credible, and may be logically sound, but are completely wrong.

Emanuele Bottazzi: Perhaps all the probabilistic approaches cannot be explanatory, since the “happen” to be wrong or right

[[BartGajderowicz|Bart Gajderowicz]]: Ideally the explanation would come from explicit knowledge. Most LLMs just don’t have that. Ensemble ML architectures may include explicit knowledge somewhere, but if the underlying processes and representations are probabilistic we reach a hard limit on explainability. Of course you can have an explanation that provides “certainty” about the answer and explanation, which is often sufficient.

Emanuele Bottazzi: I would add that ideally the explanation would come from the explicit _use_ of knowledge and principles

[[BartGajderowicz|Bart Gajderowicz]]: Do LLMs perform natural language understanding (NLU), or just processing (NLP)?

[[BartGajderowicz|Bart Gajderowicz]]: Given my definition of knowledge I’d say NLP only. Even a simple Word2vec embedding is able to identify similarity between complex objects, but I would not consider it understanding (or knowledge)

Ayya Niyyanika Bhikkhuni: “What is really true” is the underlying question when translating ancient text. The project I am working on is taking translations from humans and Generative AI and it is hoped then that people practicing according to their interpretation of the texts would tune the translations based on ‘tacit knowledge.’

Marco Monti: QUESTION: if neither LLM models nor Knowledge Graphs allow for compositionality and high contextualization of answers from a chat bot, what are the mechanisms behind the scenes of GPT X to answer so punctually and contextually ?

[[JanetSinger|Janet Singer]]: Yes, mimicry is the key characterization of what LLMs do. Parallels the 1950s it was thought that mimicry of biological behavior would inevitably lead to a structural model of living systems, and then to artificially generated life itself. See critiques by Robert Rosen.

[[GaryBergCross|Gary Berg-Cross]]: LLM based systems can learn on the job although you wouldn't call it based on experience. This has been said about the learning: "When a user interacts with an LLM-based system, the system is able to observe the user's responses and learn from them. This allows the system to improve its ability to generate responses that are relevant to the user's needs.

[[GaryBergCross|Gary Berg-Cross]]: There are a number of ways that LLM-based systems can be trained using chat responses. One common approach is to use reinforcement learning. In reinforcement learning, the system is rewarded for generating responses that are positive and helpful. This encourages the system to learn what kinds of responses are most likely to be well-received by users."

[[ToddSchneider|Todd Schneider]]: Could explain “links in OWL are not first class objects”?

Steven Wartik: Todd, a first-class object is uniquely identifiable. A reified triple is a 1st-class object.

Asiyah Yu Lin: I think the knowledge graph users who doesn't care too much about OWL thinking of data level or instance level. The ontology is really about classes. There is a blurred line between what is data and what is class.

[[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Suppose you have a model of a highway as a graph where nodes are cities and links are roads. You want to model the time it takes to get from two nodes as information directly on the link. You can do that with Neo4J but now with OWL. With OWL you need to use the design pattern where you reify the relation with a new class.

[[ToddSchneider|Todd Schneider]]: Michael, thank you for the explanation. Per your example, it could be the case that the representation (of the entities and their relations) was inadequate to support the query (i.e. with reification).
Typo “ with reification’ should be ‘Without reification).

[[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Yes. My question is how easy is it to take an OWL ontology where you have reified the relations and use graph theoretic algorithms? I don't know because I haven't used these algorithms in a long time. One thing I'm thinking about is creating an extension to OWL (I mean things like new classes and Python or SPARQL) where when you assert a new property value you have the option to create an instance of a Relation class and store data directly on that instance. That way you could treat the OWL ontology as a true graph.

[[MichaelDeBellis|Michael DeBellis]]: Often you can even ask GPT-4 to create the KIF or CycL or CLIF .. and it will

Hayden Spence: RE: Generating ontologies with LLMs: https://github.com/monarch-initiative/ontogpt

[[JanetSinger|Janet Singer]]: Here, mimicry of knowledge-driven behavior is being promoted as inevitably leading to structural models of knowledge and then to ‘emergent consciousness’. Ontologies are structural (good for modeling within their
scope); LLMs are behavioral

[[JanetSinger|Janet Singer]]: Here as in the hype cycle, not by Andrea 🙂

[[Douglas Miles|Douglas Miles]]: GPT-3 btw seems useless compared to GPT-4 on this front

Hayden Spence: From my understanding, GPT-4 is multimodal and multimodel in the sense its training is higher parameter, it incorporates more than just text data, and the actual interface is the interaction of multiple GPT models working together.

[[ToddSchneider|Todd Schneider]]: What is ‘semantic understanding’?

Hayden Spence: Is the use of established controlled vocabularies that are under license like SNOMED CT, MedDRA, ICD10/0, or standards like FHIR, and the mappings between them -- once embedded -- still restricted? At what point does transformation of information collection become its own separate from the digested information.

[[Douglas Miles|Douglas Miles]]: i don't have a question at this point.. but love this talk!

[[BartGajderowicz|Bart Gajderowicz]]: Thank you every, looking forward to the next sessions

[[JanetSinger|Janet Singer]]: Symbolic and connectionist theories of cognition are both computationalist. Leaves out 4-E embodied cognition perspective

== Resources ==
* [https://bit.ly/3rCnyC0 Video Recording]

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_04]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 04

2023-10-15T21:46:47Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Overview]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::4 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Conveners
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==

'''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''

'''Title:''' ''Fall Series Kickoff and Overview''

'''Abstract:''' The opening session of the Ontology Summit 2024 Fall Series overviews the LLM, ontology and knowledge graph landscapes, as well as introducing the participating speakers. The goal of the Series is to understand, discuss and debate the similarities, differences and overlaps across these landscapes. In addition, we will use these sessions to help to formulate the full 2024 Summit.

[https://bit.ly/3Q28U00 Slides]

[https://bit.ly/3rCnyC0 Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 4 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=04&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[AlexShkotin|Alex Shkotin]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[ToddSchneider|Todd Schneider]]
* [[MikeBennett|Mike Bennett]]
* Ayya Niyyanika Bhikkhuni
* Bill McCarthy
* Zefi Kavvadia
* [[RamSriram|Ram D Sriram]]
* Andrew McCaffrey
* Steve Wartik
* [[BartGajderowicz|Bart Gajderowicz]]
* [[MarkFox|Mark Fox]]
* Seungmin Seo
* JL Valente
* [[MichaelDeBellis|Michael DeBellis]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* Sima Yazdani
* [[JohnSowa|John Sowa]]
* [[KenBaclawski|Ken Baclawski]]
* [[RaviSharma|Ravi Sharma]]
* Sergey Rodionov
* Taj Uddin
* [[MarkRessler|Mark Ressler]]
* Asiyah Yu Lin
* Hayden Spence
* Michael Singer
* Roberta Ferrario
* Chris Novell
* Emanuele Bottazzi
* Marco Monti
* [[JanetSinger|Janet Singer]]

== Discussion ==
[[MikeBennett|Mike Bennett]]: I would not characterize a graph as a knowledge graph, even if it re-uses that label for itself. Whence the claim of 'Knowledge' in KG if not semantics? Might not be an OWL-ology of course.

[[BartGajderowicz|Bart Gajderowicz]]: An ontology is the “schema” for a knowledge graph, so it may not be designed well but there is a “schema” that defines nodes and edges in some way.

Steven Wartik: I like to distinguish between a KG and a knowledge base. A KG is a graph. It doesn't necessarily have a schema. A KB is a KG whose schema is an ontology. This is just terminology, but I find it helps my sponsors understand.

[[AlexShkotin|Alex Shkotin]]: Give me KG and I extract it's ontology.

[[KenBaclawski|Ken Baclawski]]: KGs were covered in Ontology Summit 2020. The communique has precise definitions: https://ontologforum.s3.amazonaws.com/OntologySummit2020/Communique/OntologySummit2020Communique.pdf

[[ToddSchneider|Todd Schneider]]: Knowledge =def. “facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject” (from New Oxford American Dictionary)

[[ToddSchneider|Todd Schneider]]: ‘Meaning’ is an ambiguous term.

Andrew McCaffrey: To "table" a motion means completely the opposite things in the US and the UK. :D

Ayya Niyyanika Bhikkhuni: As generative AI hallucinations become an issue, there seems a need for credibility scoring. I am about a decade out-of-the-loop, but know we were talking about this many summits ago. This is in regards to trust.

[[BartGajderowicz|Bart Gajderowicz]]: Explanations WITH hallucinations are a huge problem for LLMs. They sound credible, and may be logically sound, but are completely wrong.

Emanuele Bottazzi: Perhaps all the probabilistic approaches cannot be explanatory, since the “happen” to be wrong or right

[[BartGajderowicz|Bart Gajderowicz]]: Ideally the explanation would come from explicit knowledge. Most LLMs just don’t have that. Ensemble ML architectures may include explicit knowledge somewhere, but if the underlying processes and representations are probabilistic we reach a hard limit on explainability. Of course you can have an explanation that provides “certainty” about the answer and explanation, which is often sufficient.

Emanuele Bottazzi: I would add that ideally the explanation would come from the explicit _use_ of knowledge and principles

[[BartGajderowicz|Bart Gajderowicz]]: Do LLMs perform natural language understanding (NLU), or just processing (NLP)?

[[BartGajderowicz|Bart Gajderowicz]]: Given my definition of knowledge I’d say NLP only. Even a simple Word2vec embedding is able to identify similarity between complex objects, but I would not consider it understanding (or knowledge)

Ayya Niyyanika Bhikkhuni: “What is really true” is the underlying question when translating ancient text. The project I am working on is taking translations from humans and Generative AI and it is hoped then that people practicing according to their interpretation of the texts would tune the translations based on ‘tacit knowledge.’

Marco Monti: QUESTION: if neither LLM models nor Knowledge Graphs allow for compositionality and high contextualization of answers from a chat bot, what are the mechanisms behind the scenes of GPT X to answer so punctually and contextually ?

[[JanetSinger|Janet Singer]]: Yes, mimicry is the key characterization of what LLMs do. Parallels the 1950s it was thought that mimicry of biological behavior would inevitably lead to a structural model of living systems, and then to artificially generated life itself. See critiques by Robert Rosen.

[[GaryBergCross|Gary Berg-Cross]]: LLM based systems can learn on the job although you wouldn't call it based on experience. This has been said about the learning: "When a user interacts with an LLM-based system, the system is able to observe the user's responses and learn from them. This allows the system to improve its ability to generate responses that are relevant to the user's needs.

[[GaryBergCross|Gary Berg-Cross]]: There are a number of ways that LLM-based systems can be trained using chat responses. One common approach is to use reinforcement learning. In reinforcement learning, the system is rewarded for generating responses that are positive and helpful. This encourages the system to learn what kinds of responses are most likely to be well-received by users."

[[ToddSchneider|Todd Schneider]]: Could explain “links in OWL are not first class objects”?

Steven Wartik: Todd, a first-class object is uniquely identifiable. A reified triple is a 1st-class object.

Asiyah Yu Lin: I think the knowledge graph users who doesn't care too much about OWL thinking of data level or instance level. The ontology is really about classes. There is a blurred line between what is data and what is class.

[[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Suppose you have a model of a highway as a graph where nodes are cities and links are roads. You want to model the time it takes to get from two nodes as information directly on the link. You can do that with Neo4J but now with OWL. With OWL you need to use the design pattern where you reify the relation with a new class.

[[ToddSchneider|Todd Schneider]]: Michael, thank you for the explanation. Per your example, it could be the case that the representation (of the entities and their relations) was inadequate to support the query (i.e. with reification).
Typo “ with reification’ should be ‘Without reification).

[[MichaelDeBellis|Michael DeBellis]]: @Todd Schneider Yes. My question is how easy is it to take an OWL ontology where you have reified the relations and use graph theoretic algorithms? I don't know because I haven't used these algorithms in a long time. One thing I'm thinking about is creating an extension to OWL (I mean things like new classes and Python or SPARQL) where when you assert a new property value you have the option to create an instance of a Relation class and store data directly on that instance. That way you could treat the OWL ontology as a true graph.

[[MichaelDeBellis|Michael DeBellis]]: Often you can even ask GPT-4 to create the KIF or CycL or CLIF .. and it will

Hayden Spence: RE: Generating ontologies with LLMs: https://github.com/monarch-initiative/ontogpt

[[JanetSinger|Janet Singer]]: Here, mimicry of knowledge-driven behavior is being promoted as inevitably leading to structural models of knowledge and then to ‘emergent consciousness’. Ontologies are structural (good for modeling within their
scope); LLMs are behavioral

[[JanetSinger|Janet Singer]]: Here as in the hype cycle, not by Andrea 🙂

[[Douglas Miles|Douglas Miles]]: GPT-3 btw seems useless compared to GPT-4 on this front

Hayden Spence: From my understanding, GPT-4 is multimodal and multimodel in the sense its training is higher parameter, it incorporates more than just text data, and the actual interface is the interaction of multiple GPT models working together.

[[ToddSchneider|Todd Schneider]]: What is ‘semantic understanding’?

Hayden Spence: Is the use of established controlled vocabularies that are under license like SNOMED CT, MedDRA, ICD10/0, or standards like FHIR, and the mappings between them -- once embedded -- still restricted? At what point does transformation of information collection become its own separate from the digested information.

[[Douglas Miles|Douglas Miles]]: i don't have a question at this point.. but love this talk!

[[BartGajderowicz|Bart Gajderowicz]]: Thank you every, looking forward to the next sessions

[[JanetSinger|Janet Singer]]: Symbolic and connectionist theories of cognition are both computationalist. Leaves out 4-E embodied cognition perspective

== Resources ==

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_04]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]

ConferenceCall 2023 10 04

2023-10-15T21:46:01Z

AndreaWesterinen: /* Discussion */

{| class="wikitable" style="float:right; margin-left: 10px;" border="1" cellpadding="10"
|-
! scope="row" | Session
| [[session::Overview]]
|-
! scope="row" | Duration
| [[duration::1 hour]]
|-
! scope="row" rowspan="3" | Date/Time
| [[has date::4 Oct 2023 16:00 GMT]]
|-
| 9:00am PDT/12:00pm EDT
|-
| 4:00pm GMT/5:00pm CST
|-
! scope="row" | Conveners
| [[convener::AndreaWesterinen|Andrea Westerinen]] and [[convener::MikeBennett|Mike Bennett]]
|}

= [[OntologySummit2024|Ontology Summit 2024]] {{#show:{{PAGENAME}}|?session}} =

== Agenda ==

'''[[AndreaWesterinen|Andrea Westerinen]]''' and '''[[MikeBennett|Mike Bennett]]'''

'''Title:''' ''Fall Series Kickoff and Overview''

'''Abstract:''' The opening session of the Ontology Summit 2024 Fall Series overviews the LLM, ontology and knowledge graph landscapes, as well as introducing the participating speakers. The goal of the Series is to understand, discuss and debate the similarities, differences and overlaps across these landscapes. In addition, we will use these sessions to help to formulate the full 2024 Summit.

[https://bit.ly/3Q28U00 Slides]

[https://bit.ly/3rCnyC0 Video Recording]

== Conference Call Information ==
* Date: '''Wednesday, 4 October 2023'''
* Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
** ref: [http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=04&year=2023&hour=12&min=00&sec=0&p1=179 World Clock]
* Expected Call Duration: 1 hour
{{:OntologySummit2024/ConferenceCallInformation}}

== Participants ==
* [[AlexShkotin|Alex Shkotin]]
* [[AndreaWesterinen|Andrea Westerinen]]
* [[ToddSchneider|Todd Schneider]]
* [[MikeBennett|Mike Bennett]]
* Ayya Niyyanika Bhikkhuni
* Bill McCarthy
* Zefi Kavvadia
* [[RamSriram|Ram D Sriram]]
* Andrew McCaffrey
* Steve Wartik
* [[BartGajderowicz|Bart Gajderowicz]]
* [[MarkFox|Mark Fox]]
* Seungmin Seo
* JL Valente
* [[MichaelDeBellis|Michael DeBellis]]
* [[DouglasMiles|Douglas Miles]]
* [[GaryBergCross|Gary Berg-Cross]]
* Sima Yazdani
* [[JohnSowa|John Sowa]]
* [[KenBaclawski|Ken Baclawski]]
* [[RaviSharma|Ravi Sharma]]
* Sergey Rodionov
* Taj Uddin
* [[MarkRessler|Mark Ressler]]
* Asiyah Yu Lin
* Hayden Spence
* Michael Singer
* Roberta Ferrario
* Chris Novell
* Emanuele Bottazzi
* Marco Monti
* [[JanetSinger|Janet Singer]]

== Discussion ==
[[MikeBennett|Mike Bennett]]: I would not characterize a graph as a knowledge graph, even if it re-uses that label for itself. Whence the claim of 'Knowledge' in KG if not semantics? Might not be an OWL-ology of course.

[[BartGajderowicz|Bart Gajderowicz]]: An ontology is the “schema” for a knowledge graph, so it may not be designed well but there is a “schema” that defines nodes and edges in some way.

Steven Wartik: I like to distinguish between a KG and a knowledge base. A KG is a graph. It doesn't necessarily have a schema. A KB is a KG whose schema is an ontology. This is just terminology, but I find it helps my sponsors understand.

[[AlexShkotin|Alex Shkotin]]: Give me KG and I extract it's ontology.

[[KenBaclawski|Ken Baclawski]]: KGs were covered in Ontology Summit 2020. The communique has precise definitions: https://ontologforum.s3.amazonaws.com/OntologySummit2020/Communique/OntologySummit2020Communique.pdf

[[ToddSchneider|Todd Schneider]]: Knowledge =def. “facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject” (from New Oxford American Dictionary)

[[ToddSchneider|Todd Schneider]]: ‘Meaning’ is an ambiguous term.

Andrew McCaffrey: To "table" a motion means completely the opposite things in the US and the UK. :D

Ayya Niyyanika Bhikkhuni: As generative AI hallucinations become an issue, there seems a need for credibility scoring. I am about a decade out-of-the-loop, but know we were talking about this many summits ago. This is in regards to trust.

[[BartGajderowicz|Bart Gajderowicz]]: Explanations WITH hallucinations are a huge problem for LLMs. They sound credible, and may be logically sound, but are completely wrong.

Emanuele Bottazzi: Perhaps all the probabilistic approaches cannot be explanatory, since the “happen” to be wrong or right

[[BartGajderowicz|Bart Gajderowicz]]: Ideally the explanation would come from explicit knowledge. Most LLMs just don’t have that. Ensemble ML architectures may include explicit knowledge somewhere, but if the underlying processes and representations are probabilistic we reach a hard limit on explainability. Of course you can have an explanation that provides “certainty” about the answer and explanation, which is often sufficient.

Emanuele Bottazzi: I would add that ideally the explanation would come from the explicit _use_ of knowledge and principles

[[BartGajderowicz|Bart Gajderowicz]]: Do LLMs perform natural language understanding (NLU), or just processing (NLP)?

[[BartGajderowicz|Bart Gajderowicz]]: Given my definition of knowledge I’d say NLP only. Even a simple Word2vec embedding is able to identify similarity between complex objects, but I would not consider it understanding (or knowledge)

Ayya Niyyanika Bhikkhuni: “What is really true” is the underlying question when translating ancient text. The project I am working on is taking translations from humans and Generative AI and it is hoped then that people practicing according to their interpretation of the texts would tune the translations based on ‘tacit knowledge.’

Marco Monti: QUESTION: if neither LLM models nor Knowledge Graphs allow for compositionality and high contextualization of answers from a chat bot, what are the mechanisms behind the scenes of GPT X to answer so punctually and contextually ?

[[JanetSinger|Janet Singer]]: Yes, mimicry is the key characterization of what LLMs do. Parallels the 1950s it was thought that mimicry of biological behavior would inevitably lead to a structural model of living systems, and then to artificially generated life itself. See critiques by Robert Rosen.

[[GaryBergCross|Gary Berg-Cross]]: LLM based systems can learn on the job although you wouldn't call it based on experience. This has been said about the learning: "When a user interacts with an LLM-based system, the system is able to observe the user's responses and learn from them. This allows the system to improve its ability to generate responses that are relevant to the user's needs.

[[GaryBergCross|Gary Berg-Cross]]: There are a number of ways that LLM-based systems can be trained using chat responses. One common approach is to use reinforcement learning. In reinforcement learning, the system is rewarded for generating responses that are positive and helpful. This encourages the system to learn what kinds of responses are most likely to be well-received by users."

[[ToddSchneider|Todd Schneider]]: Could explain “links in OWL are not first class objects”?

Steven Wartik: Todd, a first-class object is uniquely identifiable. A reified triple is a 1st-class object.

Asiyah Yu Lin: I think the knowledge graph users who doesn't care too much about OWL thinking of data level or instance level. The ontology is really about classes. There is a blurred line between what is data and what is class.

[[Michael DeBellis|Michael DeBellis]]: @Todd Schneider Suppose you have a model of a highway as a graph where nodes are cities and links are roads. You want to model the time it takes to get from two nodes as information directly on the link. You can do that with Neo4J but now with OWL. With OWL you need to use the design pattern where you reify the relation with a new class.

[[ToddSchneider|Todd Schneider]]: Michael, thank you for the explanation. Per your example, it could be the case that the representation (of the entities and their relations) was inadequate to support the query (i.e. with reification).
Typo “ with reification’ should be ‘Without reification).

[[Michael DeBellis|Michael DeBellis]]: @Todd Schneider Yes. My question is how easy is it to take an OWL ontology where you have reified the relations and use graph theoretic algorithms? I don't know because I haven't used these algorithms in a long time. One thing I'm thinking about is creating an extension to OWL (I mean things like new classes and Python or SPARQL) where when you assert a new property value you have the option to create an instance of a Relation class and store data directly on that instance. That way you could treat the OWL ontology as a true graph.

[[Michael DeBellis|Michael DeBellis]]: Often you can even ask GPT-4 to create the KIF or CycL or CLIF .. and it will

Hayden Spence: RE: Generating ontologies with LLMs: https://github.com/monarch-initiative/ontogpt

[[JanetSinger|Janet Singer]]: Here, mimicry of knowledge-driven behavior is being promoted as inevitably leading to structural models of knowledge and then to ‘emergent consciousness’. Ontologies are structural (good for modeling within their
scope); LLMs are behavioral

[[JanetSinger|Janet Singer]]: Here as in the hype cycle, not by Andrea 🙂

[[Douglas Miles|Douglas Miles]]: GPT-3 btw seems useless compared to GPT-4 on this front

Hayden Spence: From my understanding, GPT-4 is multimodal and multimodel in the sense its training is higher parameter, it incorporates more than just text data, and the actual interface is the interaction of multiple GPT models working together.

[[ToddSchneider|Todd Schneider]]: What is ‘semantic understanding’?

Hayden Spence: Is the use of established controlled vocabularies that are under license like SNOMED CT, MedDRA, ICD10/0, or standards like FHIR, and the mappings between them -- once embedded -- still restricted? At what point does transformation of information collection become its own separate from the digested information.

[[Douglas Miles|Douglas Miles]]: i don't have a question at this point.. but love this talk!

[[BartGajderowicz|Bart Gajderowicz]]: Thank you every, looking forward to the next sessions

[[JanetSinger|Janet Singer]]: Symbolic and connectionist theories of cognition are both computationalist. Leaves out 4-E embodied cognition perspective

== Resources ==

== Next Meetings ==
{{#ask: [[Category:OntologySummit2024]] [[Category:Icom_conf_Conference]] [[>>ConferenceCall_2023_10_04]]
|?|?Session|mainlabel=-|order=asc|limit=3}}

[[Category:OntologySummit2024]]
[[Category:Icom_conf_Conference]]
[[Category:Occurrence| ]]