Re: need to determine what RDF is

At 11:09 30/05/2002 -0400, Peter F. Patel-Schneider wrote:

>The ``deep'' issue is that there continues to be claims that RDF
>encompasses information not encoded in RDF graphs (or in RDF-defined
>documents that can be transformed into RDF graphs).   The wording
>associated with rdfs:comment appears to be capable of supporting this view,
>although, as Pat Hayes has pointed out, it really does not.  I suppose that
>this could be considered to be just wordsmithing, but wordsmithing taking
>into account the implicit view of RDF.

One of the techniques we have found particularly useful in RDFCore has been 
reducing issues to one or more test cases.  Do you think it might be 
possible to create a test case for this issue.

I'm maybe beginning to get a glimmer of what the problem is, but I'm not 
sure.  Ramblings that are certainly technically incompetent but illustrate 
an idea follow.  I think I'm struggling to make more precise a phrase of 
Peter's "is not part of RDF".

Postulate the existence of Universal Entailment (since "RDF entailment" 
already means something else).  The Universal Entailment of an RDF graph is 
the graph which includes all the triples that a processor is entitled to 
conclude from the original graph.

I think Peter's position is that Universal Entailment contains only triples 
entailed by RDF(S) entailment.

Brian

Received on Thursday, 30 May 2002 12:36:00 UTC