- From: Piotr Kaminski <piotr@ideanest.com>
- Date: Thu, 29 Aug 2002 14:49:39 -0700
- To: Nick Bassiliades <nbassili@csd.auth.gr>, www-rdf-comments@w3.org
From: "Nick Bassiliades" <nbassili@csd.auth.gr>
> I would like to ask if it legal to transform the following set of
> triples that uses 2 rdf:type properties:
>
> A rdf:type rdfs:Class
> x rdf:type A
> x rdf:type rdfs:Class
>
> to the following set (which has just one rdf:type property):
>
> A rdf:type rdfs:Class
> A rdfs:subClassOf rdfs:Class
> x rdf:type A
No, they don't mean the same thing. If I add another triple:
y rdf:type A
then in the second model, I can also infer
y rdf:type rdfs:Class
while I can't do that in the first model. In other words, as I keep telling
my students, classification and generalization are not the same.
(Note that you could also drop the first triple from both models since it's
redundant.)
-- P.
--
Piotr Kaminski <piotr@ideanest.com> http://www.ideanest.com
It's the heart afraid of breaking that never learns to dance
Received on Thursday, 29 August 2002 17:53:39 UTC