Ontology Summit 2007: OntologySummit2007_Process proposal-1 (draft)

First proposed: Peter P. Yim / 2007.02.14

A 5-Step Process

• Convener Grouping: Summit participants ("conveners") are dispersed into 'teams' (constituencies[1]), each representing major communities within the broader ontology community space
  • we can use a survey[2] to help identify which team an individual participant is affiliated with
  • an analysis of the survey results can also provide clues as to whether we have balanced representation in different constituencies, and if not, remedial 'recruitment' actions may be taken
  • a participant can choose to be in more than one team
  • each team is encourage to elect a champion to help coordinate and track progress (this is optional, though)

• Understanding the Community Vocabulary: Each constituency will:
  • identify the context (vantage point) from which their inputs are derived; there may be needs to create further subdivisions. (For example: if "Biomedical informatics practitioners" is one such constituency, and it turns out that that the "bio-informaticists" and the medical-informaticists" don't share the same vocabularies, then they may want to subdivide and collect inputs for each of these sub-teams.
  • (each team) or each sub-team will come up with the gloss for what do they mean when they use the term "ontology"
  • identify other ontology-related terms that is in their constituency's common vocabulary,
    • and come up with the gloss for each of these terms
    • also assess the Likelihood that someone may refer to each of these terms as "ontology"
      • ( 0 = totally unlikely; 5 = some would; 9 = almost always )
  • identify conspicuous ontology (or ontology-like) artifacts[3] within their constituency
  • the above is discovered through dialog over the [ontology-summit] discussion forum, aided (where necessary) by the survey[2] and continuously summarized and synthesized and posted to the wiki to keep all participants in sync.
  • each constituency will complete and vet their glossary and list of ontology (or ontology-like) artifacts

• Constructing a categorization / typology framework[4]
  • this is done between all coneveners
  • and done in parallel with the virtual discourse described in 2. above

• Categorize
  • place terms and artifacts into the above categorization framework (by each constituency)
  • reach shared understanding on typology

• Authoring the Communique
  • draft
  • review, modify, enhance
  • adopt (with list of endorsers)
  • release / publish

Resource & References

[1] different constituencies - see under here

[2] survey - sample of a completed survey that may augment the above process - see: draft survey

[3] ontology (or ontology-like) artifacts for each constituency - see [ http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2007_Process/Draft1#nidU6N sample below]

• CYC, SUMO, DOLCE, PSL, BFO, GUM, ISO-15926
• HL7-RIM, SNOMED CT, GO, BioPAX Level 2 Ontology, ...

• • we need both name and version (as different versions of the artifact with the same name may be placed differently in the categorization framework, and hence their typology may differ

[4] categorization / typology framework - see: OntologySummit2007_FrameworksForConsideration

---

Survey questions (Draft-1)

1. Respondent information - (name, organization, e-mail, phone)

2a. Constituency affiliation

[ ] Formal ontologists community

[ ] Semantic Web communities

[ ] Concept Map community

[ ] Topic Map community

[ ] SEARCH communities

[ ] Web 2.0 community

[ ] Thesauri community

[ ] Taxonomists community

[ ] Metadata communities

[ ] XML community

[ ] Applications Development community

[ ] System Architecture Communities

[ ] Biomedical communities

2b. Expertise self-assessment

• 1 = informed layman
• 2 = practitioner
• 3 = expert

3a. Remarks (sub-community identification; vantage point; other comments)

3b. Ontology-related Vocabulary from your community

• Term
• Gloss
• Source Reference - (cite papers etc.; URL; point to section/paragrpah if possible)
• Representative Artifact - (Name, version)
• Likelihood that some may refer to this as "ontology"
  • (enter a number from 0~9: where, 0 = totally unlikely; 5 = some would; 9 = almost always )
• Comments

(repeat, say, 8 times, allowing multiple item entries on the form)

Remarks: if a participant chooses to be in more than one team (represented in more than one constituency), he/she will complete and return more than one survey.

Sample of a completed survey (on the prototype form) - (Draft-1)