| Number | 82 |
|---|---|
| Duration | 1 hour60 minute 3,600 second 0.0417 day |
| Date/Time | October 5 2020 18:00 GMT |
| 11:00am PDT/2:00pm EDT | |
| 7:00pm BST/8:00pm CEST | |
| Convener | Mike Bennett |
IAOA Semantic Web Applied Ontology (SWAO) SIG
Meetings are normally on the first Monday of the month at 2pm Eastern Time.
- We use GoToMeeting for these meetings - details at Connection Details
Agenda
- Ontology Articulation Guidelines
- New activity focusing on best practice for the semantics aspects of ontology development
- Next steps
- Housekeeping
- AoB
- Next Meeting
Proceedings
Confirm meeting time: remains as 2pm going forward.
Ontology Articulation Guidelines
See Ontolog Forum threads at present on ontology where Ed Berkmeyer, John Sowa on recognizing different kinds of ontology for different purposes
Has anyone mentioned 'engineering artifacts'?
Yes, someone did.
ToddSchneider: Mike, your audio transmission continues to break.
MB there are ontologies that are developed for an engineering solution purpose, and ontologies that are developed as concept models
These each have a purpose, but for concept models, but not necessarily an end application purpose.
Do we all subscribe to the premise that there are ontologies for applications (e.g. Semantic Web) and ontologies for reference (concept ontologies) and that these are each valid and valuable?
It follows that there are different guideliens for how to create one or the other of these.
KB: We need to be able to articulate the purpose.
Also we need guidelines for how we articulate the different kinds of purpose.
Are we presuming a common definition for what an ontology is.
MB: Yes, see Gruber.
An ontology is a formal specification of a conceptualization.
MB likes this because it is broad enough that the kinds of ontology we are trying to distinguish between, are all included within this definition.
KB dislikes it as it uses words that are themselves vague.
KB: JFS would would say that an ontology is a FOL theory, which misses the point, since theory does not necessarily have a connection to anything else.
KB you need to connect the words (or symbols? MB) to the things they are about. The words are the ontology.
TS: An ontology does nto need to be in FOL, you can use natural language.
TZ: The whole idea is to communicate something.
TS mentions the use of natural language.
TS has a problem with Gruber notion of Conceptualization.
KB sees people parroting that definition in papers all the time. But then they don't seem to use it again.
KB the Gruber definition wold also cover any computer program.
MB: a program has an ontology but is it not an ontology.
So the definition admits to some things that we would not admit as ontologies.
TS: how would that example compare with formalisms that accept of more than one interpretation.
Can we use Gruber and acknowledge that there are potential issues with it.
MB the notion of ontology is (a) it has aboutness where that aboutness is of things in the world, and has some formalism that does not admit of multiple interpretations.
KB: that works in that it rules out the interpretation of this as including computer programs.
TS it also constrains you to one interpretation.
TS correction; not that it constrains you to one interpretation, but that it constrains interpretations.
See Nicola Guarino slides on that.
MB would include higher orders of logic, mathematics also as formalisms in which you can make the formal specification of a conceptualization - would include that in ontology
all agree
Context?
We had a summit on that.
ToddSchneider: Vocabulary :: the body of words used in a particular language
ToddSchneider: But the interpretations of the 'words' in a vocabulary are left open.
MB: in terminology (AKA vocabulary in some senses), Word W refers to concept C in context X.
Meaning of a word is contextual.
https://wiki.iaoa.org/index.php/Edu:Ontology
Todd has a diagram showing the spectrum of ontologies
the page at https://wiki.iaoa.org/index.php/Edu:Term_List includes kinds of ontology.
THe ontolgoy definition page defines ontology as a count noun (correct) but definitions (1) and (2) are not of the count noun itself at all.
ToddSchneider: Ontologies by form or by function
ToddSchneider: Formalism vs. Aboutness
Todd's diagram was created in response to the very question we are addressign in this SIG so it's our first deiverable.
Todd's diagram was created in response to the very question we are addressign in this SIG so it's our first deiverable.
What else do we need to say about the distinctions between these?
TS: This is also based on Tarski model theory.
TS: There are issues with natural language since there can be ambiguity in how words are used.
Oxford English Dictionary has a new thing called Lexico that gives a partial view, for free. You don't get to see the history of the word.
Actions:
We can do a set of wiki pages on our site.
We can use the terms in the Edu term list as a starting pi=oint for our structure.
We will not be using Leo's well-known diagram for a number of reasons. We should review it howeever since it is well referenced, and to see what in that is useful for us in articulating the things we need to articulate.
MB we need to articulate the distinctions between these things before we can articulate guidelines for how one would develop these things. That's why our work is broader than simply 'Guidelines'.
Next: think about formalisms, next time. If we can't define it, at least give many examples.
ToddSchneider: Formalism :: give (something) a definite structure or shape
Next Meeting 2 November, 2pm
Attendees
- MikeBennett
- ToddSchneider
- KenBaclawski