Populating Object-Oriented Rule Engines with the Extensional Knowledge of OWL DL Reasoners

The Web Ontology Language (OWL) has become the standard for defining and sharing ontologies on the Web, based on formal and machine processable semantics. In the present work, we define a framework for importing the (extensional) knowledge about OWL instances in ObjectOriented (OO) rule engines in order to develop practical, semantic web compliant, rule-based applications. We target at domains where ontologies are used as the means for exchanging knowledge among heterogeneous environments and serve as the back-end model of rule-based applications. To this end, we import the asserted and inferred OWL axioms of Description Logic (DL) reasoners in the KB of a native rule engine in order to be matched in the body of rules, acting as constraint model. The novelty of our approach is in the fact that, instead of the trivial mapping of DL reasoners’ axioms into rule facts, we define a methodology that exploits the OO capabilities of OO rule engines. The idea is to take advantage of OO principles, such as class and attribute inheritance, and to define an OO model in such a way, so to preserve, in terms of OO relationships, the extensional semantics that have been inferred by the DL reasoner. In that way (a) the semantics of OWL ontologies are handled by sound and complete DL reasoners, and (b) instance constraints, such as instance class memberships or instance property values, are checked directly over the OO rule KB. We have implemented our methodology in the CLIPS-OWL library, an extension to the CLIPS production rule engine based on the Pellet DL reasoner that makes use of the COOL language in order to represent and check extensional OWL constraints.