This is getting into topics of considerable debate. OWLIM and others (e.g. KAON2) are datalog reasoners. Pellet and Racer use the Tableau algorithm. OWL-DL is certainly not the only way to do reasoning. Many tools provide a kind of RDFS+ that includes some of the more common inferences - symmetry, transitivity, equivalence, etc., but not full DL reasoning. The current OWL 2 working draft includes a couple of profiles using subsets of and extensions to OWL-DL. There's a draft at http://www.w3.org/TR/2008/WD-owl2-profiles-20080411/
Composer has been designed to interface with a number of these reasoning techniques. This means Composer interfaces with the reasoners and will faithfully apply the triples returned by a reasoner. Composer's task is to make editing the ontology easier, make the results of the reasoning process more transparent, combine inference engines as needed, and provide a common platform for processing different reasoning techniques. The choice of reasoning technique is largely a matter of application requirements. Add to the above mix SPARQL CONSTRUCT and Composer can be used to tailor reasoning at very fine levels. As one small example, if one does not like the tautological 'X owl:sameAs X' triples that tableau reasoners create, it is a simple matter to filter these with a SPARQLMotion script. Reasoners not directly supported by Composer can also be interfaced with using the DIG 1.x interface. This would be a good approach if one wanted to use Pellet 2, for example. There are many other refinements to reasoners that are possible with Composer's features and inference configurations. The key input to this discussion is that Composer is agnostic to specific reasoners and techniques. One needs to be aware of what the different reasoners do and then Composer can be used... to compose their results... to add process and refine reasoning results, etc. -- Scott On Jan 1, 2:58 pm, "Atanas Kiryakov" <n...@sirma.bg> wrote: > Happy new year to everyone! > > Confirming that OWLIM is not a DL reasoner. This is by design - the > inference supported by OWLIM is rule-based entailment in the style of > Datalog, OWL Horst, and OWL2 RL. This is a different type of semantics > (compared to Description Logics) - one cannot say that the one is more > powerful than the other. However, one can count that rule-based inference is > generally faster, more scalable, more manageable, and easier to understand > and tune. It is my personal believe that true DL reasoning has fairly > limited set of applications. For those one should really use Pellet or > similar reasoner. > > Regards, > Naso > > ---------------------------------------------------------- > Atanas Kiryakov > CEO of Ontotext AD,http://www.ontotext.com > Sirma Group Corp,http://www.sirma.bg > Phone: (+359 2) 8091 555; Fax: 8090 404 > ---------------------------------------------------------- > > ----- Original Message ----- > From: <schn...@fzi.de> > To: "TopBraid Composer Users" <topbraid-composer-users@googlegroups.com> > Sent: Thursday, January 01, 2009 3:58 PM > Subject: [tbc-users] Re: dl query and topbraid > > Hi! > > On 29 Dez. 2008, 17:56, Holger Knublauch <hol...@topquadrant.com> > wrote: > > To add to this: once you have configured to use an OWL DL inference > > engine (such as Pellet or OWLIM), > > Just as a side note: From looking at the OWLIM page [1], > I cannot confirm that OWLIM is a OWL DL reasoner. > OWLIM's underlying reasoning engine, called "TRREE", > is described as follows [2]: > > TRREE supports a rule laguage that is more expressive > than the one used for the definition of the RDFS semantics. > This language is almost identical with the R-Entailment > defined by Horst; the major difference is that at present > TRREE provides no support of the R-Entailment's > axiomatic triples and inconsistency rules. > > So OWLIM (or TRREE) rather seems to implement > some semantic extension of RDFS and semantic > fragment of OWL Full, which produces entailments > not covered by OWL DL. > > Cheers, > Michael > > [1] OWLIM Homepage: <http://www.ontotext.com/owlim/> > [2] TRREE Reasoner: <http://www.ontotext.com/trree/index.html> > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TopBraid Composer Users" group. To post to this group, send email to topbraid-composer-users@googlegroups.com To unsubscribe from this group, send email to topbraid-composer-users+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/topbraid-composer-users?hl=en -~----------~----~----~----~------~----~------~--~---
