Genea: Schema-Aware Mapping of Ontologies into Relational Databases
Tim Kraska, Uwe Röhm · Conference on Management of Data · 2006
Ontologies have become an important mechanism for describing and exchanging data in the semantic web, as well as in application areas such as bioinformatics or health care. This paper addresses the problem of how to eciently store and query ontology instance data using an RDBMS. Our approach automatically creates a schema-aware relational representation of the ontology using generic mapping rules. Part of the ontology reasoning, such as on subsumption relationships, is done during schema-generation and also at load-time of the instance data, so that query processing becomes faster. We implemented our approach in an OWL-mapping tool, called Genea, that can automatically create a compact relational schema and load the instance data from a given ontology written in OWL. We report on results from a quantitative and qualitative evaluation of our approach as compared to both schema-oblivious and schema-aware RDF stores from the semantic web community. Genea was found to be a generic and eective tool for ontology map