RE: rdfs:class and rdfs:resource

> Dieter
> I agree with you.
> The way I see it, every resource is EITHER an individual OR a class.

I'm not sure Dieter said that - my reading would be that you can have
instances of Class, i.e. they would be BOTH an individual AND a class.

> However, OWL FULL says that a class can be an individual.

Again I could be wrong, but although this isn't made explicit in the defs of
RDF & RDFS, there's nothing to prevent such a thing happening.

> The way I see it, that can only be true in another context.
> But OWL and RDF don't know what a context is.

Whether or not my lines above is correct, it certainly looks like much of
the confusion begins when other systems/languages are brought to the RDF(S)
table, such as the OO model or MKR/MKE contexts. This does mean that
representing RDF in these languages won't be trivial, but the APIs that
exist show it can be done. IANALogician, and don't really care too much
about what's axiomatic and what's derived, but it seems to me that as soon
as you put the blinkers on, it works fine.

Cheers,
Danny.

Received on Thursday, 8 May 2003 09:00:54 UTC