- From: Ken Baclawski <kenb@ccs.neu.edu>
- Date: Thu, 26 Apr 2001 09:34:16 -0400 (EDT)
- To: Jeen Broekstra <jbroeks@cs.vu.nl>
- cc: www-rdf-logic@w3.org, Arjohn Kampman <akam@aidministrator.nl>, conen@gmx.de
On Thu, 26 Apr 2001, Jeen Broekstra wrote:
> On Wed, 25 Apr 2001, Ken Baclawski wrote:
>
> > On Wed, 25 Apr 2001, Arjohn Kampman wrote:
> >
> > > Assume that the C is rdfs:Literal and that you're using
> > > a resource uri as a target of the property. Using (b)
> > > this would mean that the uri is a literal. To the best
> > > of my knowledge, a uri cannot be both a resource and a
> > > literal.
> >
> > No, the DAML+OIL axiomatic semantics states that literals
> > are resources. From Axiom 17:
> >
> > Ax17. (=> (Type ?r ?c) (and (Type ?r Resource) (Type ?c rdfs:Class)))
> >
> > one can conclude (by setting ?c to rdfs:Literal) that
> >
> > (=> (Type ?r rdfs:Literal) (Type ?r Resource))
>
> Maybe I misunderstand, but my impression is that this axiom
> does not hold for Literal, because membership of the set of
> literals is not expressed using an explicit type
> relationship.
That can't be true. If so, then there is a problem with this axiom:
Ax23. (=> (PropertyValue Object ?st ?o)
(and (Type ?st Statement)
(or (Type ?o Resource) (Type ?o Literal))))
How does one specify that ?o is a literal except by using
(Type ?o Literal)?
> If it _is_ to interpreted as you say, then there is a
> strange discrepancy here between RDFS (in which Literals and
> Resources are disjoint sets) and DAML+OIL.
Yes, there does appear to be a discrepancy of some kind. It would be
helpful to clarify this. As it is currently stated, Axiom 17 is very
strong. If something has any type at all, then it is a resource. It
follows that:
(<=> (Type ?r Resource) (Type ?r Thing))
Is this the intention?
Ken Baclawski
UBOT Project
College of Computer Science
Northeastern University
Received on Thursday, 26 April 2001 09:34:26 UTC