- From: Graham Klyne <graham.klyne@zoo.ox.ac.uk>
- Date: Mon, 17 Oct 2011 16:13:20 +0100
- To: Stian Soiland-Reyes <soiland-reyes@cs.manchester.ac.uk>
- CC: Paul Groth <p.t.groth@vu.nl>, Provenance Working Group WG <public-prov-wg@w3.org>
On 30/09/2011 11:06, Stian Soiland-Reyes wrote: > But what is the point of introducing prov:Time and prov:Location > classes if they have no properties and no relation to anything else in > the ontology? So they can be used as placeholder superclasses for generic applications that process and/or present provenance data without necessarily understanding the specific application-specific descriptions of time and/or location used? For example, a generic application could display rdfs:label/rdfs:description values associated with a generic time/location class without knowing how to process the specific associated values. #g -- > If it's out of scope, but we still want to say something rough about > it, we should at least introduce a proper anchoring point, like we are > (trying to) do with Role. If not then we should leave it out > completely. (Which I would not personally like, because asserting when > something happened is a quite crucial part of provenance - when its > known). > > > > For reference, here are some time/event ontologies: > > http://motools.sourceforge.net/timeline/timeline.html > http://motools.sourceforge.net/event/event.html > http://www.w3.org/TR/owl-time/ > >
Received on Monday, 17 October 2011 17:49:47 UTC