RE: Reference by description: blank nodes

Cheers Dan.

Can we cope with the situation where a B-Node may be uniquely identified not
by a single property, but by a combination of two (or more) properties?

e.g. Concepts uniquely identified by a combination of soks:prefLabel and
rdfs:isDefinedBy properties.

Al.

-----Original Message-----
From: Dan Brickley [mailto:danbri@w3.org]
Sent: 29 October 2003 17:54
To: Miles, AJ (Alistair) 
Cc: Dave Beckett (E-mail); 'public-esw-thes@w3.org'
Subject: Re: Reference by description: blank nodes



* Miles, AJ (Alistair)  <A.J.Miles@rl.ac.uk> [2003-10-29 17:25-0000]
> I found this in the Concepts and Abstract Syntax doc:
> 
> 
> 6.6 Blank Nodes
> 
> 
> The blank nodes in an RDF graph are drawn from an infinite set. This set
of
> blank nodes, the set of all RDF URI references
> <http://www.w3.org/TR/rdf-concepts/#dfn-URI-reference> and the set of all
> literals <http://www.w3.org/TR/rdf-concepts/#dfn-literal> are pairwise
> disjoint. 
> 
> Otherwise, this set of blank nodes is arbitrary. 
> 
> RDF makes no reference to any internal structure of blank nodes. Given two
> blank nodes, it is possible to determine whether or not they are the same.
> 
> 
> How is it determined whether or not two blank nodes are the same?

By being careful to distinguish the blank nodes (a data structure) from 
the things they denote. This is one reason why stopped calling them
"anonymous resources" btw. The way to figure out if two b-nodes denote 
the same resource is to do some of the reasoning licensed by the OWL
spec. In the simplest case, if you have the same property/value pair 
ascribed and the property is an owl:InverseFunctionalProperty, then the 
two nodes should denote same thing. OWL is big and complicated and there 
are doubtless many other aspects to identity reasoning via OWL that
might be relevant, but this is probably the most common case.

See http://rdfweb.org/mt/foaflog/archives/000039.html for lengthier 
explanation with a bias to how we've done this stuff in FOAF.

Dan

Received on Wednesday, 29 October 2003 13:23:20 UTC