- From: David Menendez <zednenem@psualum.com>
- Date: Tue, 04 May 2004 16:28:47 -0400
- To: Phil Dawes <pdawes@users.sf.net>
- Cc: Jeremy Carroll <jjc@hplb.hpl.hp.com>, "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>, b.fallenstein@gmx.de, pdawes@users.sourceforge.net, www-rdf-interest@w3.org
Phil Dawes writes:
>
> Jeremy Carroll writes:
> >
> > What about
> >
> > x:schnack phil:rangeIncludes y:Ghostly
> > ===>
> > _:a x:schnack _:b .
> > _:b rdf:type y:Ghostly .
> >
>
> Actually that was the inference I was trying to avoid. Unless I'm
> mistaken it effectively forces all resources that are objects in a
> triple with property x:schnak to be of rdf:type y:Ghostly.
> (as rdfs:range does).
I think Jeremy was thinking in terms of existential quantification,
rather than universal. Something like, "There is a resource Y that is
the object of x:schnack and an instance of y:Ghostly."
That is:
rdfs:range(x:schnack, y:Ghostly) ->
forall X, Y. x:schnack(X,Y) -> y:Ghostly(Y)
phil:rangeIncludes(x:schnack, y:Ghostly) ->
exists X, Y. x:schnack(X,Y) & y:Ghostly(Y)
--
David Menendez <zednenem@psualum.com> <http://www.eyrie.org/~zednenem/>
Received on Tuesday, 4 May 2004 16:29:40 UTC