On Fri, 2009-10-02 at 09:41 +0200, Martin Hepp (UniBW) wrote: > "that some properties can be expressed as text literals or > rdf:resource values". > > In Crete, we stressed the importance of staying within OWL DL > or other subsets of OWL. > > This requirement does not seem to be met by the current spec, > because using an owl:ObjectProperty with a literal is afaik > outside of OWL DL.
It's certainly suboptimal, but seems to gel with how vCard-in-RDF is used in the wild. Perhaps we should steer producers of data in one direction, while noting to consumers that historically some people have used the property differently. > http://www.w3.org/2006/vcard/Work ( rdf:type owl:Class ) > > * rdfs:subClassOf -- http://www.w3.org/2006/vcard/Address > * rdfs:subClassOf -- http://www.w3.org/2006/vcard/Label > * rdfs:subClassOf -- http://www.w3.org/2006/vcard/Tel Oh yes, ":w rdfs:subClassOf :x, :y, :z" declares :w to be a subclass of the intersection of :x, :y and :z. What we want is :w to be a subclass of their union. In RDFS/OWL this is: :w rdfs:subClassOf [ owl:unionOf (:x :y :z) ] . So for the example at hand: :Work rdfs:subClassOf [ owl:unionOf (:Address :Label :Tel) ] . -- Toby A Inkster <mailto:[email protected]> <http://tobyinkster.co.uk>
