From: Jyotishman P. <jyo...@gm...> - 2008-03-12 15:43:55
|
Hello all, I am new to this forum and wanted to learn something which might have been asked before: does the OWL API provide support for a database backend and existing triple store implementations? If yes, could you please provide me a pointer with more information on this? If not, is there an expectation to provide such a support in a future release? Thanks in advance for your reply. Regards, - Jyoti |
From: Timothy R. <tre...@st...> - 2008-03-12 16:37:15
|
Protege 3 does have its own database backend. If you look in the file menu you will see that you can convert a memory backed project to a database project. The database backend uses a Protege specific encoding of an ontology. There are no plans to build a protege 3 backend that supports existing triple store implementations. I have been thinking that such a thing would be useful and thinking about how this could be done for Protege 4. -Timothy On Wed, 2008-03-12 at 10:43 -0500, Jyotishman Pathak wrote: > Hello all, > > I am new to this forum and wanted to learn something which might have > been asked before: does the OWL API provide support for a database > backend and existing triple store implementations? If yes, could you > please provide me a pointer with more information on this? If not, is > there an expectation to provide such a support in a future release? > > Thanks in advance for your reply. > > Regards, > - Jyoti > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ Owlapi-developer mailing list Owl...@li... https://lists.sourceforge.net/lists/listinfo/owlapi-developer |
From: Matthew H. <mat...@cs...> - 2008-03-13 22:22:32
|
Hi Jyoti, > I am new to this forum and wanted to learn something which might > have been asked before: does the OWL API provide support for a > database backend and existing triple store implementations? If yes, > could you please provide me a pointer with more information on this? > If not, is there an expectation to provide such a support in a > future release? It is possible to add a new implementations of the OWLOntology interface - one could imagine that such an implementation could be based on some triple store (e.g. Sesame). In terms of in-memory triple stores then I don't think this makes sense. This is because the OWL API is generally more efficient in terms of memory usage (and load time) than a corresponding triple based API implementation. Cheers, Matthew |
From: Chuming C. <chu...@gm...> - 2008-03-16 04:22:44
|
Hi Matthew, Since axiom-centric OWL API in-memory model is more efficient than a corresponding triple based in-memory model. It is natural to think that an axiom-centric database backend would be better than a triple store? Cheers, Chuming > It is possible to add a new implementations of the OWLOntology > interface - one could imagine that such an implementation could be > based on some triple store (e.g. Sesame). In terms of in-memory > triple stores then I don't think this makes sense. This is because > the OWL API is generally more efficient in terms of memory usage (and > load time) than a corresponding triple based API implementation. > > Cheers, > > Matthew > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Owlapi-developer mailing list > Owl...@li... > https://lists.sourceforge.net/lists/listinfo/owlapi-developer > > |
From: Matthew H. <mat...@cs...> - 2008-03-18 13:13:33
|
Hi Chumming, > Since axiom-centric OWL API in-memory model is more efficient than a > corresponding triple based in-memory model. It is natural to think > that > an axiom-centric database backend would be better than a triple store? This might be a possibility, but it is difficult to say. Both of these solutions could be provided as OWL API backends though, because the OWL API doesn't commit to triples or any other concrete representation format or syntax. Personally, I am sceptical about the use of a database for storing a TBox - even with very large ontologies such as SNOMED or the NCI ontology (incidentally, the actual ontology in the NCI TBox probably accounts for about 1/5 to 1/4 of the size of the ontology -- the rest is annotations), the amount of memory required to load such ontologies isn't a vast amount considering the memory that modern machines have. Also, putting a TBox into a database generally screws reasoning since reasoners such as Pellet or FaCT++ need to "see" the whole ontology anyway, so inevitably, it gets loaded into memory. With regards to storing ABoxes in database or triple stores then I can see more sense in this if the ontology doesn't contain nominals and the ABox is very large. Cheers, Matthew > Cheers, > > Chuming >> It is possible to add a new implementations of the OWLOntology >> interface - one could imagine that such an implementation could be >> based on some triple store (e.g. Sesame). In terms of in-memory >> triple stores then I don't think this makes sense. This is because >> the OWL API is generally more efficient in terms of memory usage (and >> load time) than a corresponding triple based API implementation. >> >> Cheers, >> >> Matthew >> >> >> ------------------------------------------------------------------------- >> This SF.net email is sponsored by: Microsoft >> Defy all challenges. Microsoft(R) Visual Studio 2008. >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >> _______________________________________________ >> Owlapi-developer mailing list >> Owl...@li... >> https://lists.sourceforge.net/lists/listinfo/owlapi-developer >> >> > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Owlapi-developer mailing list > Owl...@li... > https://lists.sourceforge.net/lists/listinfo/owlapi-developer |