W3C home > Mailing lists > Public > public-rif-wg@w3.org > July 2007

Re: Issues with classification a#b, a##b

From: Jos de Bruijn <debruijn@inf.unibz.it>
Date: Tue, 31 Jul 2007 15:05:20 +0200
Message-ID: <46AF3390.60705@inf.unibz.it>
To: Michael Kifer <kifer@cs.sunysb.edu>
CC: Dave Reynolds <der@hplb.hpl.hp.com>, RIF <public-rif-wg@w3.org>

>> This still does not answer the question why RIF should introduce its own
>> ontology vocabulary.
> The reason is very simple. In this way RIF will have a standard (from the
> object-oriented point of view) minimalistic theory of class hierarchies.
> Any other theory that has its own subclass mechanisms (like RDF) can be
> hooked up to RIF's theory using the simple rules like
> ?p # ?t :- ?p[rdf:type -> ?t] .
> ?p ## ?t :- ?p[rdfs:subClassOf -> ?t] .
> Thus, the different theories that will be imported into different RIF
> modules will be able to interoperate and integrate their class hierarchies
> through the RIF's simple model.

There are already two standard Semantic Web languages for the purpose of
interoperating and integrating ontological information (including class
hierarchies), namely RDFS and OWL.  I argue that we should use these
languages for expressing and integrating class information, rather than
inventing a new one.
I know that RDF Schema has some features which make the language overly
complex.  However, through syntactical restriction things can be made a
lot easier (e.g. [1], which includes all the standard uses of the RDFS
ontology vocabulary).

Best, Jos

[1] Sergio Muñoz, Jorge Pérez, Claudio Gutiérrez: Minimal Deductive
Systems for RDF. ESWC 2007: 53-67.

> 	--michael  
Jos de Bruijn            debruijn@inf.unibz.it
As far as the laws of mathematics refer to
reality, they are not certain; and as far as
they are certain, they do not refer to
  -- Albert Einstein

Received on Tuesday, 31 July 2007 13:05:31 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:47:46 UTC