RE: rdfs:domain for OO class properties

I had the same problem with dc:title and dc:creator like predicates.
(dc = dublincore)

musicbrainz used the dc: namespace to define its author relation. when I
did a RDF scheme out of this, I did a domain definition that caused
problems:

<!-- !!!!!!!! WWWRRRROOOONNNNNGGGGGGGG -->
<rdf:Property rdf:about="&dc;creator"
	 rdfs:label="creator">
	<rdfs:domain rdf:resource="&mm;Album"/>
	<rdfs:domain rdf:resource="&mm;Track"/>
	<rdfs:range rdf:resource="&mm;Artist"/>
</rdf:Property>

Dave Reynolds read it and suggested the following solution in
http://lists.w3.org/Archives/Public/www-rdf-interest/2003Apr/0299.html
Dave:
"Second, your declaration of DC:title is saying that it applies to
objects which
are type mm:Album AND type mm:Artist AND type mm:Track - whereas I
assume you
really mean OR. That effect could be achieved in RDFS by creating a new
class
mm:HasTitle and saying that mm:Album etc are sub-classes of it."

RDF Schema Validator - Validate RDF triples

> -----Original Message-----
> From: www-rdf-interest-request@w3.org 
> [mailto:www-rdf-interest-request@w3.org] On Behalf Of 
> Francesco Cannistrà
> Sent: Tuesday, July 15, 2003 10:54 AM
> To: Pedro Assis in Oporto; www-rdf-interest@w3.org
> Subject: Re: rdfs:domain for OO class properties
> 
> 
> 
> Hi
> 
> > Assuming that the following interpretation of the property domain is
> > correct, what should be the right procedure (formally!?) to 
> describe in
> > RDF a property which can be applied to several distinct classes?
> >
> > The context is: The RDF properties domain (rdfs:domain) are globally
> > defined and the conjunction of domains, i.e. more than one 
> statement for a
> > property domain, leads to the class domain that is of both, (in the
> > case of 2 domain statements) types.
> >
> > Well, one solution is to create different properties, i.e. 
> with different
> > names, to describe the same behavior (hum, don't like 
> that). Another one
> > is to declare the domain as a collection of alternate 
> classes, i.e. all
> > the members of such collection (not sure about this one). 
> But, necessarily
> > this issue has arise many times before, so it must have a 
> proper solution.
> 
> Another one (the rignt one I think) is to use OWL that is 
> going to be born
> just to let do things like this.
> 
> Regards
> 
>

Received on Tuesday, 15 July 2003 05:21:03 UTC