Re: Coordination between RDF(S) and DAML+OIL

From: Peter F. Patel-Schneider ([email protected])
Date: 06/24/01


So far this is only one person's effort.  It might be better to see if any
changes are made this week.

peter


From: Dan Brickley <[email protected]>
Subject: Re: Coordination between RDF(S) and DAML+OIL
Date: Sun, 24 Jun 2001 22:40:59 +0100 (BST)

> 
> Thanks, that's a handy summary. Mind if I pass this on to the relevant
> W3C groups? (Semantic Web Coordination Group; RDF Core WG; RDF Interest
> Group...).
> 
> Regarding RDF Schema, I intend to open up discussion of some RDFS issues
> within the RDF Core WG shortly. Will keep you posted...
> 
> cheers,
> 
> Dan
> 
> On Sun, 24 Jun 2001, Peter F. Patel-Schneider wrote:
> 
> > As promised, here is my initial list of coordination points between RDF(S)
> > and DAML+OIL.
> > 
> > peter
> > 
> > 
> > 	Coordination between RDF(S) and DAML+OIL
> > 
> > - what does DAML+OIL depend on from RDF(S)
> >   - RDF - basic triple model
> >   - RDFS - basic RDFS classes - CLASS, RELATION, ...
> > 	 - class organization - subclass, subproperty(?)
> > 	 - structuring relationships - domain, range, ...
> > 
> > - what does DAML+OIL not use at all
> >   - RDF - reification - not needed, not understood
> > 	- containers - wrong properties for our purposes
> >   - RDFS - meta-class organization - not needed
> > 	 - ...
> > 
> > - what changes does DAML+OIL require
> >   - RDFS - multiple domains - allow, with conjunctive reading
> > 	 - multiple ranges - change from disjunctive to conjunctive reading
> > 	 - subclass can be reflexive - 
> > 
> > - what areas are problematic
> >        - fit in the middle
> >        - have problems
> >        - are missing from RDF(S)
> >   - datatypes - currently in DAML+OIL
> > 	      - should be in RDF(S)
> >   - simple class organization - currently in RDFS
> > 			      - should be ???
> >   - reification - lots of problems
> > 		- ...
> >   - containers - need semantic justification
> > 	       - perhaps move elsewhere ?
> >   - domains and ranges - problem with multiple domains and ranges
> >   - metaclasses and extensibility - ?
> >   - no structure for information in RDF(S)
> > 
> > - other issues
> >   - RDF and RDFS are not good layers
> >     - both provide simple stuff, but both also provide suspect stuff
> >       - RDF - triples / reification
> >       - RDFS - frames / global ranges, but no local restrictions
> >     - RDF gives meaning to all syntax
> >       - makes it hard to define extensions
> >   - what is a URI? - syntax and semantics
> > 
> > 
> 


This archive was generated by hypermail 2.1.4 : 04/02/02 EST