W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > April to June 2009

Re: Potential text for time-permitting features in F&R

From: Birte Glimm <birte.glimm@comlab.ox.ac.uk>
Date: Tue, 30 Jun 2009 06:21:10 +0100
Message-ID: <492f2b0b0906292221h2b3683ben4ee5cd499d699ca4@mail.gmail.com>
To: "Seaborne, Andy" <andy.seaborne@hp.com>
Cc: Ivan Herman <ivan@w3.org>, SPARQL Working Group <public-rdf-dawg@w3.org>
> While we loosely refer to "graph matching" it is a misnomer.  Sec 12.6 defines the extension framework in terms of answers to basic graph pattern.  There is no implied completing the graph with triples - which is impossible in some cases anyway; there is only returning answers.
;-) You are right.

> Do you have a concrete example for SPARQL where they (direct and RDF semantics) differ?

For the OWL WG conformance tests, all tests are classified as either
direct or RDF based semantics and AFAIK there are some OWL WG tests
that have, for the same test, different results for the two semantics,
but I don't have any off the top of my head since I usually just run
the direct semantics tests. These tests check, for example, for
entailment between ontologies and could most likely be rephrased as
SPARQL queries. If it is important, I can try and find a concrete
example.
Birte


>        Andy
>
>> -----Original Message-----
>> From: public-rdf-dawg-request@w3.org [mailto:public-rdf-dawg-
>> request@w3.org] On Behalf Of Birte Glimm
>> Sent: 29 June 2009 16:44
>> To: Ivan Herman
>> Cc: SPARQL Working Group
>> Subject: Re: Potential text for time-permitting features in F&R
>>
>> Ivan,
>> sorry for answering late. My only problem with RDF based semantics is
>> that it is not clear to me how this can be done properly with logical
>> entailment rather than graph matching. In my understanding, you could
>> either use SPARQL as it is now (with graph matching) to query OWL
>> ontologies or for something closer to the entailment semantics, you
>> would have to complete the graph with possibly an infinite number of
>> new triples, which is not really feasible. It could work for the RL
>> profile, so that might be something to start with. I am not opposed to
>> mention both semantics since it is time permitting anyway, but the
>> direct semantics is achievable IMHO, whereas I have doubts about the
>> RDF based one, apart for OWL RL.
>> Birte
>



-- 
Dr. Birte Glimm, Room 306
Computing Laboratory
Parks Road
Oxford
OX1 3QD
United Kingdom
+44 (0)1865 283529
Received on Tuesday, 30 June 2009 05:21:51 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:15:39 GMT