Extracting Satisfiability-Preserving Modules from the OWL RL Ontology for Efficient Reasoning

Abstract

Reasoning on large ontologies has been identified as an important challenge. Concept satisfiability is one of the core reasoning tasks in ontology engineering. To improve the efficiency of concept satisfiability checking for OWL RL ontologies, we propose an approach for extracting modules from the OWL RL ontology. Our approach is based on the idea that for a given concept being checked, not all of the ontology elements are related to the checking process. By transforming the OWL RL axioms into first-order constraints and analyzing the dependencies among the constraints, we give a strategy for deleting the irrelevant elements; thus, a simplified ontology that preserves the original checking result can be obtained. The experimental results obtained for a set of real-world ontologies show that our method can effectively reduce the reasoning time, and the efficiency is increased by 1.15 times to 9.08 times.