Friday, July 16, 2004

RDF Querying going to the DAWGs

The most recent DAWG face-to-face raised some interesting issues with respect to using XQuery for querying RDF. Jim Hendler has a response:

"You show RDFS/OWL/Rule query langauges as somehow being more easy inXquery, but again I think that is because you are assuming these things will be kept in their RDF/XML documents, or in APIs that respect the "boundaries" of those. I already see many applications moving towards multiontologies w/linking, and that seems to me to argue that we simply don't know yet which of these models are better."

The original proposal suggests that we're going to need a query language for OWL, Rules and RDF, which probably won't happen and it's suggested without proof. It also suggests that because RDF can be serialized in XML that it has something in common with XQuery, which is untrue. The standard RDF/XML serialization can have multiple forms of the same RDF graph. The same RDF query will work across different RDF/XML serializations, because it is operating on the same data model, this isn't true for XQuery.

It reminded me of the recent anti-XQuery article, "If You Liked SQL,You'll Love XQUERY".

Fabian is saying that the relational model was a simplification of graph theory. In this respect relational theory and RDF have much in common, much more in common than XML.

At a syntactic level, query languages like RDQL, iTQL and other SQL-like RDF query languages are leveraging off a legacy of SQL, Datalog and other similar languages. This is something that XQuery lacks as well. Do we really want FLWOR and Conditional Expressions in our query language?

Fabian also mentions NULLs, a continual pet peeve of the anti-SQL crowd, it's good to see XQuery avoids this. Something that I hope an RDF query language avoids as well.

Interestingly, Don Chamberlin's XQuery tutorial is quoted both by Fabian and in Jeff's proposal.

Andrae is also blogging some of this as well, "Jumping the gun".
Post a Comment