Actions

Ontolog Forum

Session Interoperability
Duration 1 hour
Date/Time 13 May 2026 16:00 GMT
9:00am PDT/12:00pm EDT
5:00pm BST/6:00pm CST
Convener Todd Schneider

Ontology Summit 2026 Interoperability

Conference Call Information

Discussion

12:06:17 deddy : What is relationship between OAGi & OMG… another standards body?

12:09:46 deddy : Fabulous AG entertainment on YouTube

  • Laura Farms
  • Larson Farms
  • Millennial Farmer

They're into the fields for planting.

Incredibly complex computer driven machines.

12:13:40 deddy : Good luck eliminating cash outflows when combines START at $500,000+

12:20:45 TS : Can AI be used in creating or revising standards in the context in the value they may engender?

12:23:52 Ram Sriram : Recent report on Standards and Innovation: https://www.ansi.org/standards-news/all-news/5-11-26-standards-are-a-meaningful-driver-of-innovation

12:24:54 Ram Sriram : Paper written more than two decades ago on Standards and Innovation: https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=821473

12:25:53 deddy : "Philadelphia’s Philosopher Mechanics: A History of the Franklin Institute 1824 - 1865" by Bruce Sinclair

Three efforts:

  1. Water power
  2. Steamship boiler explosions
  3. Weights & measured standards

12:27:58 Ram Sriram : You can contact Jacob Collard at NIST for publications related to this: https://www.nist.gov/programs-projects/linguistic-and-ai-support-collective-standards-development

12:28:03 Michael DeBellis : This is not what I was expecting in a talk about interoperability. When I think about Interoperability, the last thing I think of is Porter. I see Interoperability as broadly two dimensions:

  1. Communication interoperability: message and event busses like RabbitMQ, Kafka, connectors that resolve issues between calling an API in Java vs. one in Python. Integrating two systems where both use knowledge graphs but one uses Neo4J and another uses OWL/RDF Standards like MCP.
  2. Semantic Interoperability. Resolving the fact that within the same company several groups have concepts like Customer but Customer mean different things for Marketing vs. Fulfillment, vs. Catalog and Shopping Cart teams. IT in the past usually just does this ad hoc. I worked with HL7 for a while and at the time, trying to define a common object model was greeted with wooden stakes and garlic but decades later they do it with HL7 FHIR.
  • deddy : Folks at the top will need to allocate resources. Top level trained on Porter.
    • Michael DeBellis : Assuming that is true, and that Porter is more than just standard MBA business babble, you could say the same thing about any problem: building better reasoners, supporting Big Data, etc. it all has to be justified at a business level so not sure why it matters more for Interoperability.
  • Ram Sriram : What is the current State of Art on using FHIR for Semantic Interoperability?
    • Michael DeBellis : Don't know. I think it has made some surprisingly good traction but haven't been working with it.

12:28:38 TS : How can we discern if AI responses have incorporated standards?

12:30:53 John Sowa : When the technology is developing so fast, it is too soon to propose and enforce any standards.

12:31:52 John Sowa : So far, he hasn't told us what his technology is and what does it do.

  • Michael DeBellis : 👍
  • Terry : 👍

12:34:50 Michael DeBellis : No on "Ontologies need to be Perfect". The best way to derail a project is to assume you need to get it right and perfect the first time. That's why we don't use the Waterfall model. Because you just aren't going to get it right the first time and the best way you know how you are wrong is to deploy something and get feedback from the users. And in spite of what some people say: change is constant in business so even if you get the ontology perfect, it's going to change so you need an Agile process.

  • deddy : You assume there’s no upward error correction portion of SDLC?
  • Gary Berg-Cross : To enable interoperability you might have to uncover hidden assumptions, clarify value trade-offs, and contextual constraints.

12:38:04 TS : The SCOPE Model for Interoperability Assessment can help is understanding and distinguishing context.

12:38:47 Gary Berg-Cross : I found Guizzardi, Giancarlo. "Ontology, ontologies and the “I” of FAIR." Data Intelligence 2.1-2 (2020): 181-191. useful for interoperability in a FAIR/ontology context. Giancarlo would have been a good speaker to recruit.

12:39:53 deddy : I’m holding my breath on when the drawing tools standardize on arrowhead meaning(s). < sigh >

  • janet singer : 😀

12:40:44 janet singer : This approach provides the broader context for thinking through the relationships between IT standards, interoperability and ontologies.

12:40:48 Mike Bennett : In all OMG modeling standards, each line and line ending has a defined meaning. Unlike the more casual drawing tools. Model semantics is the key.

12:41:19 deddy : OMG embraces UML…?

  • Mike Bennett : OMG wrote UML
  • deddy : Ah, ha! So it has to be a standard. < sigh >
  • deddy : My involvement with OMG pre-dates UML… so I missed it.

I’ve been exposed to, but never worked in a UML environment.

12:42:49 Mike Bennett : The advantage of a standard is precisely that you know what the different arrowheads mean because they have standard meanings.

12:45:17 janet singer : ‘Model semantics’ is a proposed standard to be used (de jure, de facto).

12:45:59 deddy : Does Archimate use UML?

12:50:54 janet singer : UML was created by bringing together OOAD approaches by Booch, Rumbaugh and Jacobson then adopted by OMG

  • deddy : thx for genealogy

12:51:37 John Sowa : logic is fundamental.

12:52:06 John Sowa : UML and every other kind of diagram can be specified in logic.

  • David Eddy : But systems are built on political power… with some logic on the side.
  • TS : John, communication is fundamental; Logic is an aid.
    • John Sowa : Then it Is possible to translate that logic to anything in any notation -- including any and every natural language.
  • David Eddy : .. and as always ignoring the existence / importance of unnatural language? The language used inside the operational systems

12:52:50 Terry : Chart 81 - Continuum doesn't address (un) predictability

12:54:48 Gary Berg-Cross : It is better to say that ontologies provide the semantics for metadata and thus understanding data.

  • David Eddy : You can say it.. doesn’t make it accurate or useful.

12:54:58 janet singer : Wirfs-Brock’s role-responsibility approach should have been incorporated as well, but it was challenging to incorporate (as I recall)

12:55:48 Gary Berg-Cross : Reference ontologies still matter.

  • David Eddy : How long does it take me to wrap an ontology around the regulatory report my summer intern is assigned to?

12:57:23 Gary Berg-Cross : Environment Ontology (ENVO)is relevant for Ag with biomes, environmental features, and habitat descriptors. But there are some more specific.

12:58:12 Gary Berg-Cross : Candidate ontologies and vocabularies include: AGROVOC

  • Scope: multilingual controlled vocabulary for agriculture, food, fisheries, forestry, environment.
  • Use: indexing, metadata, semantic search; widely used by FAO and research repositories.

Crop Ontology (CO)

  • Scope: standardized trait, phenotype, and trial descriptors for many crops (maize, rice, wheat, cassava, etc.).
  • Use: phenotype data annotation, breeding databases, germplasm characterization.

Plant Ontology (PO)

  • Scope: plant structures and growth/developmental stages across species.
  • Use: linking phenotypic data, gene expression, and anatomical terms.

12:59:31 Gary Berg-Cross : For agronomic practices, experimental designs, management activities. we have Agronomy Ontology (AgrO). But harmonizing all of these is a big issue especially for interop.

13:00:31 David Eddy : Ahhhhh… “boundaries”… an ambiguous topic

13:01:14 Gary Berg-Cross : BTW, the AgroPortal / AGROVOC has a linked KG

  • Focus: semantic integration of agricultural vocabularies (AGROVOC + others).
  • Uses: metadata enrichment, term linking, powering semantic search across datasets.

13:03:00 Gary Berg-Cross : Wageningen University / WUR was mentioned. They seem to have some knowledge graphs that may be relevant (e.g., AgroLD). IS OAGI using or supporting things like this?

13:04:12 Gary Berg-Cross : FoodKG / FoodOn also provides some integrations of food products, supply chain, safety records linked to provenance and taxonomy.

13:06:18 Gary Berg-Cross : To ensure interoperability effective KGs for AG would need to combine resources like Crop Ontology, Plant Ontology, ENVO, AGROVOC, SAREF4AGRI, MIAPPE mappings .

13:07:34 David Eddy : A Minnesota family farm… starting planting: Larson Farms https://www.youtube.com/watch?v=tqYJkQZ_B_w

13:08:24 John Sowa : The methods for mapping a dialog in English to and from Common Logic are precise. Most importantly they exist and work -- successfully.

  • David Eddy : Does this Common Logic work to map from operational systems, written in decidedly unnatural languages & meanings back into natural language?
    • janet singer : Seeing languages as standardized codes is how one can see that ‘unnatural’ language and natural language are more alike than different, so their issues need to be addressed together

13:15:47 John Sowa : I am NOT suggesting that anybody learn or use Common Logic. I am saying that they can communicate precisely in their own native language.

  • David Eddy : I’m searching for ability to map “back” from operational systems to whatever…

13:18:11 David Eddy : Duly note there is no Wikipedia page for IBM’S AD/CYCLE metadata effort in late ‘80s.

  • Gary Berg-Cross : David, I hope that we have advanced from the 1980's on these issues.

13:19:49 janet singer : If all languages (human, machine, etc.) were appreciated as standardized codes for the agents in various communities, then common logic could be seen in the Rosetta Stone role

Resources

Previous Meetings

 Session
ConferenceCall 2026 05 06Interoperability
ConferenceCall 2026 04 29Interoperability
ConferenceCall 2026 04 22Synthesis
... further results

Next Meetings

 Session
ConferenceCall 2026 05 20Education
ConferenceCall 2026 05 27Education
ConferenceCall 2026 06 03Cognition
... further results
Retrieved from "https://ontologforum.com/index.php?title=ConferenceCall_2026_05_13&oldid=5582"