Expressive approximations in DL-lite ontologies

  • Authors:
  • Elena Botoeva;Diego Calvanese;Mariano Rodriguez-Muro

  • Affiliations:
  • KRDB Research Centre, Free University of Bozen-Bolzano, Italy;KRDB Research Centre, Free University of Bozen-Bolzano, Italy;KRDB Research Centre, Free University of Bozen-Bolzano, Italy

  • Venue:
  • AIMSA'10 Proceedings of the 14th international conference on Artificial intelligence: methodology, systems, and applications
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Ontology based data access (OBDA) is concerned with providing access to typically very large data sources through a mediating conceptual layer that allows one to improve answers to user queries by taking into account domain knowledge. In the context of OBDA applications, an important issue is that of reusing existing domain ontologies. However, such ontologies are often formulated in expressive languages, which are incompatible with the requirements of efficiently accessing large amounts of data. Approximation of such ontologies by means of less expressive ones has been proposed as a possible solution to this problem. In this work we present our approach to semantic (as opposed to syntactic) approximation of OWL 2 TBoxes by means of TBoxes in DL-LiteA. The point of interest in DL-LiteA approximations is capturing entailments involving chains of existential role restrictions, which can play an essential role in query answering. The presence of TBox assertions involving existential chains affects query answering by enriching the number of obtained rewritings, and hence allows us to cope better with incomplete information about object and data properties. We provide an approximation algorithm and show its soundness and completeness. We also discuss the implementation of the algorithm.