Hello Baran,
Kind regards, Lorenz > > I think statements like > > On Fri, 25 Aug 2017 11:52:46 +0200, Lorenz Buehmann > <buehm...@informatik.uni-leipzig.de> wrote: > >> .... Inferencing and querying are totally different >> things. So why are you thinking about refactoring the whole project? > > or in next posting > >> Again, why do you compare those two Jena mechanisms? What is the >> expected outcome? > > are generally spoken confining snd mistakable for Jena-users making > thoughts about a proper design. > > Assume i have following dev-scenario: > > A Jena-app with InfModel + Rules-List -> output RDF -> TDB/Fuseki -> A > Query-UI > > (Query UI contains a lot of queries which users activate with a mouse > click and get responses presented in the same UI.) > > Now you i can try to add rules to my rules-list so that i can > formulate some of the queries of my Query-UI in a more leightweight > way with better performance. > > (Or vice versa i change some of my queries so that i can delete some > rules of my rules-list which is not so interesting.) The question here was to compare rules vs. SPARQL CONSTRUCT queries to modify an existing dataset. Indeed, the combination of both might be more powerful, although it remains open what part of the semantics to cover by which mechanism. Nevertheless, it's totally use case dependent - what kind of rules, how many rules, size of the dataset, is the data volatile (forward chaining vs. backward chaining), what kind of queries, and so on and so forth. Good luck with the project. > baran > -- Lorenz Bühmann AKSW group, University of Leipzig Group: http://aksw.org - semantic web research center
