- From: Patrick Stickler <patrick.stickler@nokia.com>
- Date: Wed, 06 Feb 2002 20:03:40 +0200
- To: Pat Hayes <phayes@ai.uwf.edu>, Jeremy Carroll <jjc@hplb.hpl.hp.com>
- CC: RDF Core <w3c-rdfcore-wg@w3.org>
On 2002-02-06 18:04, "ext Pat Hayes" <phayes@ai.uwf.edu> wrote: >> My question is about the lack of an explicit triple giving the dType, this >> was one of my problem cases in the TDL model theory. > > The whole point of the 'ranging' semantic condition is to allow > <range> info to entail <dtype> info, so that if the 'local' idiom > works then the 'global' one will as well (though you have to do some > extra inference, of course). It works for all the cases that use > bnodes, but it doesn't work very well for the inline-literal case. > > Thats also why one needs to use a separate namespace (or some > equivalent trick) to stop having too many possible leaks from > rdfs:range to rdf:type. If we only used local typing there wouldnt be > any real need for that. > > Pat If we went with the revised global idiom that is a derivative of the local idiom with rdf:type optional, but with the bNode I.e. Bob ex:age _:1 . _:1 rdf:value "30" . Then can we also get away with rdf:type rather than rdf:dtype? If so, then I think that's another point in favor of the convergence proposal, since we don't have to add any new vocabulary to RDF at all. Yes? No? Cheers, Patrick -- Patrick Stickler Phone: +358 50 483 9453 Senior Research Scientist Fax: +358 7180 35409 Nokia Research Center Email: patrick.stickler@nokia.com
Received on Wednesday, 6 February 2002 13:02:27 UTC