SEM: owl:Class vs. rdfs:Class was: Re: WOWG: agenda Aug 15 telecon

Jim Hendler wrote:
>
> To the best of my recollection, the discussion of the precise
> differences between daml:class and rdfs:class, which seems to carry
> over into our current decision-making and issues, was discussed in
> the DAML joint committee, and not really in this WG -- I've looked
> through the archives and seen many references to owl:class being
> different than rdfs:class, but I cannot find a message that precisely
> describes the differences -- can someone either point the WG to such
> a message (or import one from joint-committee WG archives) or write a
> summary.  From our own discussions, it is not clear to me what is
> being offered as a reason not to simply drop owl:class and use
> rdfs:class (or just make them equivalent if we resolve the open issue
> to have everything be owl:)

I think that this is a central question for the Semantics of OWL:

Is owl:Class defined as according to the OWL model theory (or whatever
semantic definition is chosen) different (at all) than rdfs:Class as defined
according to the RDF model theory?

If the answer to this question is YES, then we need owl:Class, if the answer
to this question is NO, then we don't _need_ owl:Class. If this question
cannot be answered (and I am fearing that it cannot in a straightforward and
consistent fashion), then we ought to pack it up and go home until we can
get a straight answer to this basic question.

Jonathan

Received on Thursday, 15 August 2002 10:48:16 UTC