RE: RDF Graph questions

> On Fri, 24 May 2002, Jeremy Carroll wrote:
>
> > <rdf:RDF>
> >  <rdf:Description rdf:about="http://example.org/"/>
> > </rdf:RDF>
>
> FWIW, I always mentally parse this as syntactic longhand for:
>
>  <rdf:RDF>
>   <rdfs:Resource rdf:about="http://example.org/"/>
>  </rdf:RDF>
>
> 'rdf:Description' is a way of saying "there is a resource and
> it has the
> following (URI, properties...)". But it doesn't explicitly assert the
> 'there exists a thing and it is of rdf:type rdfs:Resource
> bit. Which is
> fine, cos everything's a resource, so its a pretty vacuous
> assertion to
> make.

This differs from Dave's position I fear.

If we take the rdfs schema closure of the graph corresponding to the
given RDF/XML file then it contains two triples

rdf:type rdf:type rdf:Property .
http://example.org/ rdf:type rdfs:Resource .

(according to DanB).

This can be formalized within our current framework by saying that an
RDF Graph is a set of nodes and a set of arcs, where each arc is a
triple [ subj, pred, obj ] where subj and obj are in the set of nodes.

Dave's position can be formalized as saying that an RDF Graph is a set
of arcs, and the nodes in the graph are defined as the set of nodes in
the arcs. Isolated nodes that do not partake in any triples are
prohibited.

I currently agree with Dave, but fear this looks like an issue.

Jeremy

Received on Friday, 24 May 2002 08:50:10 UTC