Conjunctive Query Optimization in OWL2-DL

Conjunctive query answering is becoming a very important task on the Semantic Web as the adoption of SPARQL query language increases. There is considerable work done in the area of optimizing conjunctive query answering for RDF and OWL2-DL ontologies, in the latter case namely for queries without undistinguished variables. However, there has not been much emphasis on how to handle queries with both distinguished and undistinguished variables efficiently. In this paper, we present a novel algorithm for answering conjunctive queries over OWL2-DL ontologies with undistinguished variables efficiently. These optimizations are implemented in the Pellet query engine and our experimental evaluation shows that the new optimization improves the query performance significantly.