3

I was thinking about reasoning in ontologies.

What are the the differences between the inferecen capabilities between DL reasoners (like Hermit, Pellet, etc) and Prolog?

I mean, let us suppose that I have a Knowledge base containing an ontology (general axioms) and facts. What can I infer from this with Prolog and what can I infer with DL reasoners?

It would be interesting if you could provide references as well.

Best regards

Zaratruta
  • 2,097
  • 2
  • 20
  • 26
  • 3
    why don't you read about the basics of both first? I mean, the most obvious difference is *closed world* vs. *open world assumption*, isn't it? Read about NaF as well. By the way, again there is already research out there that for examples bridges the gap between OWL and Prolog programs, e.g. conversion from OWL to Prolog( e.g. Thea2 etc) There are also fancy reasoners that even do support Tableau reasoning in Prolog, e.g. Trill. And many other papers, approaches, tools, etc. – UninformedUser Oct 29 '18 at 09:24
  • I have read the basics of both. I know that they are different regarding the closed world assumption. But I don't understand if they can do the same think regardless this point. – Zaratruta Oct 29 '18 at 16:33
  • I'd also add that they are based on different tractable fragments of FOL. Roughly speaking, Prolog is based on Horn clauses, while OWL is based on description logics, which in essence are the so-called _guarded fragments_ of FOL. Obviously, intersections, combinations etc. should be possible (as AKSW has pointed out). BTW, SWRL is Datalog in essence, so OWL RL more or less too... – Stanislav Kralin Oct 30 '18 at 06:16

0 Answers0