Re: Comments on SPARQL Query Language for RDF (21 July 2005 version)

Jim Hendler <hendler@cs.umd.edu> on Fri, 2 Sep 2005 wrote:

 > Dan Connolly wrote (in response to Nokia's comment):
 >
 > > > On Thu, 2005-09-01 at 13:12 -0400, Art.Barstow@nokia.com wrote:
 > > >
 > > > 1) This document does not discuss in any way the
 > > > *semantics* of the query language. We would like to see
 > > > a more formal definition of queries (and their results)
 > > > in terms of RDF semantics (right now, the query language
 > > > seems to treat RDF graphs as merely data structures from
 > > > which something can be extracted). Why would SPARQL now
 > > > ignore RDF's model theory when one was created through
 > > > a sizeable effort?

I believe that the current SPARQL does not ignore RDF MT, and it has  
been carefully crafted to be in sync with it. On the other hand, this  
document is not a document on the semantics of SPARQL. I believe that  
a specific new document fully describing the semantics of SPARQL and  
its properties is needed, in particular by emphasising its grounding  
on RDF MT, on standard DB query languages theory, and classical logic  
model theory.

 > > >  3) It does not seem possible to extend SPARQL to be
 > > >  used with OWL (primarily, perhaps, because of comment
 > > >  #1 above).
 > >
 > > A number of WG members (UMD, Agfa) are succesfully using SPARQL  
with OWL.
 >
 > UMD is not willing to be listed as claiming that we are successfully
 > using SPARQL with OWL.
 >    We are implementing numerous OWL based tools and making them
 > coexist with OWL, but that is quite a different thing.   I do not see
 > that SPARQL gets us anything more significant in the OWL space than
 > the RDF graph queries for data or from the specific graphs that OWL
 > represents.  However, that is a long way from "successfully using
 > SPARQL with OWL" and, as the SPARQL spec has evolved we have
 > considerably scaled back our expectations of the what SPARQL will do
 > for us.
 >   We have continued to support the design of SPARQL because we do a
 > significant amount of work with respect to RDF triplestores and we
 > very much want to do distributed triple store linking, but we do not
 > consider SPARQL to be very useful for OWL as per Nokia's comments.

I disagree.
There are various conceivable ways to let RDF, SPARQL, and OWL  
interoperate - even in an elegant, efficient, and doable way.
We have formally defined in [1] a framework where it is possible to  
query RDF(S) graphs via SPARQL given a background ontology in  
arbitrary fragments of FOL (e.g., DLs or OWL-DL/Lite). We found that  
it is possible to use a plain DL reasoner to encode (efficiently) the  
framework when DL/OWL-DL ontologies are involved. We even proved that  
the limitations imposed in the W3C semantic definition of the  
relation between OWL-DL/Lite and RDF(S) are not necessary, and we  
propose a direct relation between them.
This was possible since we did in advance a thorough analysis of the  
logical foundations of RDF(S), that sheds a new (brilliant) light on  
the formal foundations of RDF(S) and SPARQL, and on their  
interoperability and usefulness also in presence of OWL-DL ontologies.

cheers
--e.

[1] <http://www.inf.unibz.it/krdb/w3c/rdf-sparql-semantics.pdf>

Enrico Franconi                  - franconi@inf.unibz.it
Free University of Bozen-Bolzano - http://www.inf.unibz.it/~franconi/
Faculty of Computer Science      - Phone: (+39) 0471-016-120
I-39100 Bozen-Bolzano BZ, Italy  - Fax:   (+39) 0471-016-129

Received on Saturday, 3 September 2005 22:39:43 UTC