Re: Historical events

I agree, it would be nice to create a type to capture those properties that provide other information than the boundaries of a time period.

If Event is limited to the narrow definition as Roger has noted, “‘an event happening at a certain time and location, such as a concert, lecture or festival.’ “Its properties—composer, performer, organizer, audience, doorTime, workPerformed, review etc.—clearly adhere to this narrow definition.” — then it would be very useful to have something like schema.org/Occurence<http://schema.org/Occurence> for other types of events.


On May 30, 2018, at 1:15 PM, Vicki Tardif <vtardif@google.com<mailto:vtardif@google.com>> wrote:

I think we need to be cautious not to stretch semantics too far, especially based on whether or not certain properties could be used to express an idea. If time boundaries are all a thing needs to be an Event, then we do not need schema.org/Person<http://schema.org/Person> as we can use startDate and endDate for birthDate and deathDate.

Looking at Wikidata, if you follow the hierarchy, WWII is an occurrence, which loosely maps to event, but the Bronze Age is a time period, which seems about right to me.

Getting back to the original question, ISO 8601 can be used to give the rough boundaries of a time period, but this use case (and probably others) requires adding properties to the time period and you can't do that with an ISO 8601 string. It might be nice to create a type to capture those properties.

- Vicki


On Wed, May 30, 2018 at 3:01 PM Thad Guidry <thadguidry@gmail.com<mailto:thadguidry@gmail.com>> wrote:
Vicki,

I'm just saying it is not a real necessity because Event taken as something within a time boundary, covers it all.  But if folks want to classify sub-Types of Events...sure, why not.

I myself get very far just by using Thing's properties and the Event "about" property to classify sub-Types of Events.  Which tells us we are doing things right on those levels at least, being able to model long-tail domains even with our existing classes and properties...although I understand for Roger, Allison, and others, it might be too much work for them.

Sure, I'm fine with introducing a new class, where it could have correlationEvent as a new property, as in the following resource link.

Just having these basic classes added into a new Schema.org<http://Schema.org> extension could go a long way
http://resource.geosciml.org/ontology/timescale/gts#classes

-Thad

On Wed, May 30, 2018 at 1:20 PM Vicki Tardif <vtardif@google.com<mailto:vtardif@google.com>> wrote:
To Roger's point, I think we should distinguish between events and ages or eras. I think it is fine to call WWII an "event", but the Bronze Age or the 1960s are more timeframes when events happened rather than the events themselves.

- Vicki


....................................................
Allison Muri
Department of English

Arts 418
University of Saskatchewan
Saskatoon, SK, Canada
ph: 306.966.5503