Abstract
Abstract. Ontology Based Data Access (OBDA) has drawn considerable attention from the OWL and RDF communities. In OBDA, instance data is accessed by means of mappings, which state the relationship between the data in a data source (e.g., an RDBMSs) and the vocabulary of an ontology. In this paper we present Quest, a new system for OBDA focused on fast and efficient reasoning with large ontologies and large volumes of data. Quest provides SPARQL query answering with OWL 2 QL/RDFS entailments and can function as a traditional OWL reasoner/triple store, or as a mediator, located on-top of a legacy data source linked to the ontology by means of mappings. In such configuration all data remains in the data source and is only accessed at run-time. Quest uses query rewriting techniques as the inference mechanism in both modes. In this paper we describe the architecture of Quest, and the optimization techniques it currently implements.