|Duration||1 hour60 minute |
|Date/Time||March 01 2021 19:00 GMT|
|11:00am PST/2:00pm EST|
|7:00pm GMT/8:00pm CET|
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
- Latest post for the IAOA website
- Ontology Articulation Guidelines
- Presentation: Andrea Westerinen on Events ontology
- Presentation: Mike Bennett on Ontology Styles and TLO usage
- Next Meeting
No new blog post yet this month - awaiting closure on these issues (today's meeting)
This is on the discussion about TLOs
Slides also related to IoF discussion last week on event and process etc.
Trips ontology (from Rochester)
Andrea Westerinen: https://www.cs.rochester.edu/research/trips/lexicon/browse-ont-lex.html
In the notion of Event, is there a Subject?
There is an Agent. Need not be a Subject (e.g. climate event in Ethiopia). There are agents who cause / experience things.
TS: Subject = what the Event is about
AW: That is sometimes called the Theme. What the event is about it the event itself (hurricane, meeting eetc.) so it stands alone per Davidson view in linguistic theory of events. Has arguments, things people do or are done to them etc.
TS: Many people would use the word Subject for those e.g. subect = some meeting.
AW: That is the Happening, Occurrence or State (the 'What is this about?')
Versus Subject Predicate Object (where the Event is the Subject in a triple)
AW: approaching it in terms of the way people think
Whys: Goal etc v Causation (2 kinds of Why)
ToddSchneider: A web drawing tool, https://excalidraw.com
Andrea Westerinen: http://www.linguist.org/History-of-Events.pdf
Vendler 1967 (above)
Deals with the refinements of kinds of Event based on the presence of otherwise of the various properties in Andrea's slide.
Also reflected in DOLCE
Using RDF* to do probabilistic analysis or reasoning
Eventuality covers both Event and State (also in that paper) i.e. Stative and Eventive
Also looking at VerbNet (from Colorado) - we should look at that.
If Failure in there? AW has a model based on this, took Failure out, also has Success etc.
What is the relationship between Event and Observation?
- using measurement as an event
TS: What is the truth maker for Event?
Ascribe probability to that agent?
- you ascribe provenance to the information about the event.
Uses booleans for things like if an event is planned etc.
Robert rovetto: Some additional concepts on events are (a) event calculus,(b) simple event model, (c) concepts by Galton and Mizoguchi, (d) an unpublished 2012 paper in a univ workshop on events I wrote attempting a taxonomy of events
Is a process the same as an Event?
No - process is a sequence of events.
KB: Events can have a complex structure e.g. Ontology Summit v Summit Session
So you have a mereology
KB: That makes it difficult to talk about hasTime for a thing
AW: Not at all - the Summit (complex event) has a start and end, the sessions have their own times.
KB: the meanings of the times in those 2 cases are quite different.
AW: The collection of meetings has its own time
Robert rovetto: (e) there’s also I think a recurrent event ontology design pattern
ToddSchneider: If a 'process' is a sequence of 'Events', is a single 'Event' a 'Process'?
The boolean (property of owl:thing) covers the distinctions like, planned, future, possibility etc. as well as generic for punning
AW not keen to introduce sub classes corresponding to these as it affects future modeling possibilities.
TS: Process v Event
Process as a sequence of events?
AW: or States
TS: If I have a single event is that a process?
Process has a start and end (sub properties of Process) + well-defined steps
Also things that are not process, such as a riot, or global warming
MB: the word Process may refer to several things, one of which is our notion of business process. Global warming is a kind of process but it is not a kind of one of those.
TS: There is a more general concept of which Natural Process and Business Process would be a kind.
MB: Nice to find common abstractions, but did not want to assume that every use of a word will necessarily point to a common concept.
See Oxford US English definition
has several distinct definitions.
Business process ('Process') has goal, end point etc.
Commonality - some change is happening
MB: Maybe Process is how a change happens
A Sequence of Events can be a natural process too.
RR: Would not limit Business Processes to those having Goals. Would not say that those processes that have goals are business processes. Non human processes also have a teleological element. Have artificial events and processes.
MB: Unpack teleology between conscious (agent with a goal) versus more broadly function driven
RR: The literature backs up that (latter) use of these terms.
AW: Sequencing: there is a hasNext
Slides can be shared with the IoF?
Yes- AW will clean them up.
How would this approach play into the IoF commitment to BFO?
TS: It provides another perspective to the participants.
Many of them are not familiar with DOLCE or the options that exist in TLO.
KB: What is the intention behind this Event ontology?
AW: Using for narrative analysis. Trying to put together how people think about cause and effect and how they explain their lives in terms of narrative events.
Would you ever use this beyond that?
AW has used this same ontology in 2 different programs, in causal explanations and in logistics work. Has helped focus the ontology on particular problems.
KB has worked on such programs also (and logistics) and would see this Event ontology as being helpful.
KB: Are you looking to make this a standard?
AW: Not really; just putting out ideas. Resonates with non-ontologists. Talking with SMEs, others.
KB Would be nice if this could be re-used
AW: Happy to support reuse of this.
AW will write some papers on this ontology, but also describe what you can do with it - not simply an intellectual exercise.
KB: The database community is also interested in events.
AW happy to take this anywhere.
Deals with the verbs side of the world, having previously focused on the nouns side of things. e.g. shipping, ordering, maintenance events. Lets you think about who is participating, what gets done etc.
Who will write the blog post?
Are we ready to write one yet?
AW would like to take some of today's feedback (e.g. Todd o change, Robert's points) and revise the slides.
KB: Post can simply say you are working on something.
RR would love to brainstorm more on events
MB: Maybe we should continue with the Events conversation (e.g. MB can do a counter-proposed Events conceptual ontology) for this thread of thinking; may help frame our thoughts on TLO more generally.
TS: How well can these things be merged with BFO and if not, why?
TS It would be interesting to understand how this approach to Event might conflict with BFO (for IoF).
RR would encourage looking at Galton and Mizoguchi’s approach to events also - Event Calculus
AW has worked with the Event Calculus
Propose a call in 2 weeks to continue this discussion
All seem to think this is a good idea.
Can we invite one or two others who would be able to contribute to this conversation
For example some of the IoF folks.
Next call: 2 weeks time, invite selectively, keep it small. Agenda is specifically Events. Also make people aware of other TLOs.
Also talk bot DOLCE, gUFO etc. on this.
MB: to arrange this in 2 weeks. 15 March Same time same dial-ins.
14:00 EDT (not EST – clocks will have changed in North America)
These meetings are pinned to US prevailing time.
- Mike Bennett
- Ken Baclawski
- Andrea Westerinen
- Todd Schneider
- Robert Rovetto