On 12 May 2009, at 11:26, John Goodwin wrote:
If I could be sure that my SPARQL query environment would use
OWL reasoning to fill in the gaps in its knowledge, then I'd
be more inclined to publish OWL data. But right now there is
a dearth of high-quality, widely-used OWL implementations.
Well there are implementations of SPARQL-DL (e.g. Pellet) but most
SPARQL endpoints I know do tend to just do RDFS reasoning (if that).
Maybe demand will drive more implementations?
I know there is a new SPARQL working group - is there much talk of OWL
inference + SPARQL going on there?
Yes, there is. That doesn't directly address the problem of how you
run a OWL-DL reasoner over even medium sized datasets (say, one
gigatriple). As far as I know that is an unsolved problem, though I
could well be out of date in that front.
My understanding is that the issues around SPARQL/OWL are to do with
how you get consistent results across OWL reasoners, and how you
represent certain types of entailment in SPARQL results.
- Steve
--
Steve Harris
Garlik Limited, 2 Sheen Road, Richmond, TW9 1AE, UK
+44(0)20 8973 2465 http://www.garlik.com/
Registered in England and Wales 535 7233 VAT # 849 0517 11
Registered office: Thames House, Portsmouth Road, Esher, Surrey, KT10
9AD
