OWL Query Answering Based on Query Extension

The paper presents an approach for optimizing query answering algorithms that are based on approximate instance retrieval. We consider SPARQL instance queries over OWL ontologies and use the OWL 2 Direct Semantics entailment regime of SPARQL for their evaluation. Approximate query answering algorithms are based on the creation of two sets; the set of certain or known query answers and the set of possible query answers, which require checks to determine whether they are real answers. Typically, it is expensive to check the possible answers hence our goal in this paper is to reduce the number of possible answers returned by approximate reasoning algorithms. We present an approach for using schema knowledge from the terminology (TBox) to optimize the evaluation of SPARQL instance queries. We proceed by transforming the query into a set of assertions (ABox). We then show how the TBox and this (small) query ABox can be used to build an equivalent query where the additional query atoms can be used for reducing the set of possible mappings for query variables.