SWAO:Conference Call 20220114

Number	95
Duration	1 hour60 minute 3,600 second 0.0417 day
Date/Time	January 14 2022 19:00 GMT
	11:00am PST/2:00pm EST
	7:00pm GMT/8:00pm CET
Convener	Andrea Westerinen

IAOA Semantic Web Applied Ontology (SWAO) SIG

THIS MEETING TOOK PLACE ON FRIDAY 14 JANUARY 2022. (Meetings have moved to the second Friday of the month at 2pm Eastern Time.)

• We use Zoom for these meetings - details at Connection Details

Agenda

• Actions from last meeting
  • RobertR liaison status with the IAOA Education Committee still has not occurred since no meetings have yet been held
• Discussion/presentation of RobertR's ontology definitions spreadsheet
• Discussion of the MIRO article
• Continued discussion of ontology and knowledge graph definitions from the December SWAO meeting
  • Posted in this document
• Housekeeping and next meeting
• AoB

Proceedings

• Preliminary discussion about the antifraud ontology announced on the Ontology Forum distribution list
  • GAO Antifraud Resource is based on GAO’s Fraud Ontology — "a rigorous classification of fraud schemes affecting federal programs and operations—serves as the backbone for understanding, evaluating, and measuring all aspects of federal fraud schemes, including their participants, mechanisms, and impacts"
    • The site provides a user friendly, web-based platform for interacting with the model and identifying resources to support fraud risk management
    • This project was overviewed by Leia Dickerson in the 2021 Ontology Summit series
    • Question was raised whether this could be more integrated with cyber security

• For the next 6 months or so, SWAO will need another individual to run the meetings (AndreaW has several other commitments, as does MikeB)
  • Either ToddS or KenB will run the sessions for the coming months
  • AndreaW updated the SWAO meeting connection details to reflect KenB's Zoom information

• Discussion of upper ontologies migrated to a discussion of IOF (Industry Ontology Foundry)
  • There is concern that all upper ontologies could be rejected due to issues and systemic bias for a single one (such as BFO)
  • BFO issues include the requirement for realism and the lack of multiple inheritance
  • Committing to an upper ontology means committing to its conceptual distinctions, formal semantics, philosophical worldviews, metaphysical biases, etc., and thus claims about how the world is
    • These need to be evaluated for relevance, agreement with potential users and developers, complexity, and consistency with the beliefs and desired world views (if any) of the ontology to be developed
  • Is valuable to understand where BFO is used and where it is used but its issues are "worked around"
  • The majority of the following insights on IOF were provided by ToddS:
    • Development principles for IOF are in harmony with OntoCommons
    • IOF is less concerned about BFO's requirement of realism and more concerned with 'practical application' for physics and mathematics for manufacturing
    • IOF Core to be released soon, with BFO formally embedded
    • A release where BFO is "hidden" has also been discussed as a possibility

• Discussion of the MIRO article
  • MIRO is the acronym for "Minimum Information for Reporting an Ontology"
  • Provides a set of guidelines to facilitate consistency and completeness when documenting published ontologies
  • There are 7 areas for documentation, and 3-11 guidelines within each area (34 guidelines total)
    • 28 guidelines are MUSTs, 3 SHOULDs and 3 OPTIONALs
    • Topic areas are 'basics', motivation, scope and dev community, knowledge acquisition, content, change mgmt, and testing/QA
    • It was noted that this seems heavy but can be valuable for very large, heavily referenced/reused ontologies
  • As part of the discussion, 'incremental development of semantics' was discussed
    • This is addressed in a forthcoming paper in the Journal of Applied Ontology, "Issues in Incrementally Adding Better Semantics to Knowledge Graphs" by Gary Berg-Cross
    • The theory is to add data to an ontology/knowledge graph and then fix the ontology in order to support it
    • This is as opposed to a top-down/upper ontology definitional approach
    • The underlying definition of the concepts is in TypeScript

• Discussion of ontology concepts by RobertR
  • Need alternate wording than upper/top-level/... ontology since these imply direction and specific architectures, as opposed to intent
    • "Core" implies "center" with other ontologies extending from it or layering around it
    • "Upper" implies that other ontologies are below it and that there are levels of gradation
    • "Top-level" vs "foundational" imply that other ontologies are "below"/"above" them
      • For example, in computer science, hierarchies are typically drawn with the most general at the top, whereas in biology, the most general is at the bottom
    • "Conceptual" ontology could be either broader or narrower
  • "General", "generic", "common" ontology may be more appropriate names
  • Take-aways:
    • Important to have definitions for the various types of ontologies
    • Should explore the possible architectures of linked ontologies with varying degrees of generality
    • Does it matter if an ontology is only categorical or abstract, or also includes instances?
    • Important to understand how the world is perceived/partitioned, and put the conceptual archetypes within that structure
      • For example, "prescribed vs described", "occurrent vs endurant", "dependent/relative/mediating", "concrete vs abstract" , ...
    • Important to define these concepts and then assess whether they are needed when developing an ontology
    • Need to distinguish between "truth makers" (what something is/semantic building blocks) and the data that one has about the thing
      • Also need to understand the quality of the data

Next Meeting

The next meeting will occur on February 11, 2022, at 19:00 GMT/2pm EST.

Attendees

• Ken Baclawski
• Mike Bennett
• Robert Rovetto
• Todd Schneider
• Andrea Westerinen

Previous Meetings

• Conference Call 20211210
• Conference Call 20211011
• Conference Call 20210712

... further results