Re: [ALL] RDF Data Access, XQuery, rules

I noted in today's agenda:
[[
6.  Comments to DAWG on RDF Data Access, XQuery, Rules

     [ALL] RDF Data Access, XQuery, rules
     http://lists.w3.org/Archives/Public/public-swbp-wg/2004Sep/0049.html

     Any volunteers for a second reviewer?
]]


I've also looked at:

http://www.w3.org/TR/2004/WD-rdf-dawg-uc-20040802/

and saw nothing there that I thought needed this WG's attention,
although I did make a comment about literal matching ... (particularly 
worrying about I18N issues)

http://lists.w3.org/Archives/Public/public-rdf-dawg-comments/2004Sep/0007.html


In terms of Dan's specific questions:

1.
 > Anybody have experience with XQuery/RDF integration to share?

No, I was happy with the downplaying of XQuery in this doc.
I am also happy to see that they are intending to work with XPath 
Functions and Operators, which should mean that they don't drift too far 
from the XQuery view of the world.

Gary's
 >   If either the above answers is no, or we don’t know,
 >  then we shall not
 >  offer any views on this item.

and the lack of response suggests not offering a view on this item 
(perhaps an explicit 'no comment').


2.
 > Any rules/query integration experience to share? Thoughts

I found the sections in the use cases adequate on this:
[[

4.6 Additional Semantic Information
It should be possible for knowledge encoded in other semantic 
languages—for example: RDFS, OWL, and SWRL—to affect the results of 
queries executed against RDF graphs.

4.6a Additional Semantic Information (variant)
It should be possible for a query to indicate that the answers should 
take into account knowledge encoded in RDF semantic extensions such as 
RDFS, OWL, etc.

Status: Pending.

]]

Yes - the Jena team allow this, as described in:
http://www.hpl.hp.com/techreports/2003/HPL-2003-146.html

We have our rule systems set up as graph-to-graph transforms, doing e.g. 
RDFS or OWL inference.
The RDQL query lanaguage can then be used to query either the base graph 
(without inference) or the virtual graph (with inference). Both are useful.

Particularly problem with OWL is that the virtual graph is infinite, 
that can sometimes be surprising.

Gary says:
 >   From a development and deployment point of view I think it
 > makes a lot of sense to be able to use a single query language
 > across different inferencing system to access instance data.
 > The choice is then to match inferencing capability to the
 > expressive that the problem demands.

I agree. However, the practical difficulties with OWL, since it is on 
the limit of terminating and complete tractability, mean that in 
practice there will be more implementation variability than in DAWG 
without inference.

Jeremy

Received on Thursday, 30 September 2004 10:05:25 UTC