- From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
- Date: Sun, 14 Oct 2001 07:33:17 -0400
- To: GK@ninebynine.org
- Cc: www-rdf-interest@w3.org
From: Graham Klyne <GK@ninebynine.org> Subject: Re: a new way of thinking about RDF and RDF Schema Date: Sun, 14 Oct 2001 11:44:28 +0100 > At 03:52 PM 10/12/01 -0400, Peter F. Patel-Schneider wrote: > > >3/ Interpretations > > > >An interpretation I is a four-tuple > > < IR, IEXT, ICEXT, IS > > > > >where IR is a non-empty set, called resources > > IEXT <= powerset ( IR x (IR u DV) ) > > ICEXT : IR -> powerset ( IR u DV ) > > IS : U -> IR > > > >and IS(rdf:type) in ICEXT(IS(rdf:Property)) > > > > ICEXT(IS(rdf:Description)) = IR > > Did you mean rdf:Resource here? Nope. This is a significantly different way of looking at RDF, and one of the changes would be to make rdf:Description be the top of the ``food chain''. In the rdfs section, rdfs:Resource is set up as a ``co-top''. > > ICEXT(IS(rdf:Property)) <= IR > > > > if < x , y > in IEXT, y in ICEXT(IS(rdf:type)), and < y , z > in IEXT > > then x in ICEXT ( z ) > > Er, the members of IEXT are sets of pairs, not bare pairs, no? Again, this is a major change of viewpoint (but just of viewpoint, not of content). IEXT here is bare pairs. To get to the situation in Pat Hayes's model theory, you have to group two pairs together to get a ``relationship''. This new model theory is in many ways more similar to one of the model theories for conceptual graphs. I have a slightly-modified version of my model theory, which I have not distributed yet, that gives a hint at the difference here. An RDF interpretation can be turned into one of Pat Hayes's interpretations by taking each pair of tuples <x,p> and <p,z> in IEXT where p is in P [think of p as an ``instance'' of a property] and replacing them with <x,z> in IEXT(r) for each r such that p in ICEXT(r), then adding <x,c> in IEXT(IS(rdf:type)) for each x in ICEXT(c). What I should do is to change the names IEXT and ICEXT to something different to indicate their different role. > #g peter
Received on Sunday, 14 October 2001 07:33:27 UTC