On Wednesday 03 September 2008 08:01:03 you wrote: > Hi, you're obviously ahead of me in testing the RDF stores out there, so > would you mind describing why you're looking at virtuoso instead of jena > now? I gather it's mostly for speed--
Indeed! > are you seeing good speedups for > the queries that you -have- figured out how to run? I haven't actually run any queries on Virtuoso yet, and basically, I have been just a victim of massive advertising from Kingsley and Ted ;-) Most queries are not problematic in Jena. We have found, however, that the order of the statements of the pattern matters a lot in Jena, one must always, manually see to that the most restrictive terms go first, and that is a hard problem to solve when the queries are generated rather than hand-written. I don't know if Openlink has done anything to solve this problem, or if there are other things that could help in that respect, but I'm here to find out. The other problem, which is now urgent to us, is that "simple search engine" problem. Since we are stuffing all data into the model several times to have it indexed, the length of the literals seem to cause us trouble. We are working on several fronts with this problem, one solution involves Virtuoso, some does not. We are using Jena as a SPARQL Endpoint, but we are also using SPARQL Update. With the current architecture, it should be straightforward to switch back and forth, and that's what we intend to do. Kind regards Kjetil Kjernsmo -- Senior Knowledge Engineer Direct: +47 6783 1136 | Mobile: +47 986 48 234 Email: kjetil.kjern...@computas.com Web: http://www.computas.com/ | SHARE YOUR KNOWLEDGE | Computas AS Vollsveien 9, PO Box 482, N-1327 Lysaker | Phone:+47 6783 1000 | Fax:+47 6783 1001
