- From: Enrico Franconi <franconi@inf.unibz.it>
- Date: Tue, 7 Oct 2003 09:52:34 +0200
- To: "Raphael Volz" <rvo@aifb.uni-karlsruhe.de>
- Cc: <www-rdf-logic@w3.org>
From: "Raphael Volz" <rvo@aifb.uni-karlsruhe.de> > > > I believe that you - generally - can NOT implement OWL Lite > > > with rule-based reasoners. > > > > It is not only a belief, you can formally prove it! No sub-boolean DL > > (e.g., FL-) with full axioms, and no DL with full booleans (e.g. ALC) > > without axioms can be ever encoded in a polynomial language (e.g., a > > rule-based reasoner), as a matter of fact. > > I would like to stay with my belief, you may choose to do some > preprocessing and compile things and later use your polynomial language > to do the rest. I don't get your point. A proof is stronger than a belief :-) > > > What you seems to be possible > > > is using Disjunctive Logic Programs, but our proof that an > > > appropriate translation is indeed correct is not quite finished > > > yet. Empirically, we pass all tests in the OWL Test Suite and > > > other well-known DL tests. > > > > Still, DLP in its full power is below PSPACE, so again no reasonable DL > > (with a complexity coming either from full axioms or from the booleans) > > can be ever be encoded in it. This is the case of OWL-lite which is > > EXPTIME-hard. So, again, I don't understand what are you really doing. > > > Boris Motik is working on the proof, we will see if he can prove it or > not. This can't be, unless you are not using disjunctive logic programming. Did you get my complexity argument? What is your answer to this? Is there something I am missing? Please let me understand fully your proposal (and all the proposal which use a rule based approach to reasoning with OWL). cheers --e. Enrico Franconi - franconi@inf.unibz.it Free University of Bozen-Bolzano - http://www.inf.unibz.it/~franconi/ Faculty of Computer Science - Phone: (+39) 0471-315-642 I-39100 Bozen-Bolzano BZ, Italy - Fax: (+39) 0471-315-649
Received on Tuesday, 7 October 2003 03:53:19 UTC