- From: Geoff Chappell <geoff@sover.net>
- Date: Sun, 6 Mar 2005 17:06:15 -0500
- To: "'ben syverson'" <w3@likn.org>, <semantic-web@w3.org>
> -----Original Message----- > From: semantic-web-request@w3.org [mailto:semantic-web-request@w3.org] On > Behalf Of ben syverson > Sent: Sunday, March 06, 2005 4:28 PM > To: semantic-web@w3.org > Subject: Re: Style question > > > > On Mar 6, 2005, at 7:41 AM, Geoff Chappell wrote: > > [...] > > i.e. becca is a person with green eyes. The value of eyeColor for any > > person > > must be one of the defined EyeColors (green or brown). > > Actually, I was shooting simply for: Becca is a person with green and > brown eyes (brown in the center, green at the edges, specifically, but > lets assume the user doesn't want to get that specific). In that case, maybe you just want to say something like this: <?xml version="1.0"?> <rdf:RDF xmlns:ex="http://likn.org/#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:owl="http://www.w3.org/2002/07/owl#"> <ex:person rdf:about="http://likn.org/#becca"> <ex:eyeColor rdf:resource="http://likn.org/#green"/> <ex:eyeColor rdf:resource="http://likn.org/#brown"/> </ex:person> </rdf:RDF> I'd personally stay away from bags - they don't really carry much meaning. For example, you can't infer that a relation between a resource and a bag holds between the resource and the members of the bag. > But this raises an interesting question: I'd like to keep the ontology > in a separate file, so that if a constraint on "person" is updated, I > don't have to update every person RDF. Is that outsourcing too much > information from the main files people will be looking at (the RDFs)? It often makes sense to have class and instance data separate. If you do, it's probably a good idea to either make the schema retrievable at its uri or point to it via owl:imports or rdfs:seeAlso in the instance data. > Also, what do I do with a line like: > <likn:instanceOf rdf:resource="http://likn.org/#person" /> > Is there a way to define through OWL that anything that is a > likn:instanceOf is also an owl instance of the target resource's class? You could make likn:instanceOf a rdfs:subProperty of rdf:type. But why not just use rdf:type? Is likn:instanceOf really a specialization of rdf:type in any way or is it just another name for the same thing? > And how would I tie the two files together? Or should I be doing this > differently? The ontology should be general enough that individuals > such as "Becca" don't show up in the ontology file unless they have > specific restraints... > > I'm concerned that rdf:type by itself is too vague. For example: > <dc:title>Music Box Theater</dc:title> > <rdf:type rdf:resource="http://likn.org/#theater" /> > Does this imply that the Music Box is a type of theater, when in fact > it's an instance of Theater? It says that the resource is an instance of theater. It does not say that it is a subclass of theater (which I think is what you mean by "is a type of theater") - for that you'd say: <rdfs:subClassOf rdf:resource="http://likn.org/#theater" /> > - ben Geoff
Received on Sunday, 6 March 2005 22:11:32 UTC