1

I'm working with graphDB (ontotext data base graphs) and I would like reproduce in graphDB (.pie file) the behaviour of Pellet reasoner in Protege. In this way, I want to get in graphDB the same behaviour that I have in Protege. I have implemented rules in GraphDB but I'm not sure how reproduce the Pellet behaviour. I have read about OWL-DL and DL semantics (Pellet is build on that language) but I'm not find rules that implements and how reproduce it in GraphDB.

Any idea to reproduce that behaviour??

thanks

Sergio
  • 81
  • 9
  • There exist standard "rulesets": http://graphdb.ontotext.com/free/devhub/inference.html#standard-rule-sets. Select e.g. "OWL-Max" when creating your repository. – Stanislav Kralin Nov 02 '17 at 19:29
  • Thanks for your answer! ...please, another question, only to be sure... OWL-MAX corresponds to "builtin_Rules.pie" file?? what ruleset is "builtin_RulesPublishing.pie"??? – Sergio Nov 03 '17 at 07:57
  • 1
    ...but, is not "OWL-MAX" an implementation of OWL-Lite (less expressive than OWL-DL)?? ( http://graphdb.ontotext.com/documentation/free/reasoning.html#predefined-rulesets ) – Sergio Nov 03 '17 at 08:22

1 Answers1

2

You can not reproduce in GraphDB precisely the same behavior like in Pellet or the opposite direction either. Pellet and GraphDB (formerly known as OWLIM) use two absolutely different reasoning approaches. GraphDB by design is a database limited to only tractable fragment so the loading time, where all inference happens, stays predictable.

The diagram bellow gives a high-level summary how the different semantic fragments overlap.

GraphDB semantic expressivity compared to OWL fragments

vassil_momtchev
  • 1,173
  • 5
  • 11