OntoQL: An Alternative to Semantic Web Query Languages

Abstract

Ontologies are used in several application domains for representing knowledge. The defined approaches differ according to the type of addressed ontology (conceptual or linguistic) and to the used ontology model (e.g. OWL or PLIB). Several languages have been proposed to manipulate ontologies and their instances, especially in the Semantic Web domain. However these languages are often specific to a given ontology model, they focus on conceptual ontologies and they are not compatible with database exploitation languages. We address these three problems in this paper by proposing the OntoQL language. This language has three main original characteristics: (1) OntoQL is based on a core ontology model composed of the shared constructors of ontology models. This core ontology model can be extended by the language itself, (2) OntoQL queries can be expressed with different natural languages features using the linguistic layer of an ontology, and (3) OntoQL is fully compatible with SQL enabling a smooth integration between SQL queries of classical database applications and ontological queries. As a theoretical validation of this language, we present the algebra of operators that sets up its formal semantics. On the operational side, we describe the implementation of OntoQL on the OntoDB database and we illustrate the interest of this language by reporting several applications where this language has been extensively used and proved powerful.