On the discovery of subsumption relations for the alignment of ontologies

  • Authors:

  • Vassilis Spiliopoulos;George A. Vouros;Vangelis Karkaletsis

  • Affiliations:

  • AI Lab, Information and Communication Systems Engineering Department, University of the Aegean, Karlovassi, Samos 83 200, Greece;AI Lab, Information and Communication Systems Engineering Department, University of the Aegean, Karlovassi, Samos 83 200, Greece;Institution of Informatics and Telecommunications, NCSR "Demokritos", Athens, Greece

  • Venue:
  • Web Semantics: Science, Services and Agents on the World Wide Web
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

For the effective alignment of ontologies, the subsumption mappings between the elements of the source and target ontologies play a crucial role, as much as equivalence mappings do. This paper presents the ''Classification-Based Learning of Subsumption Relations'' (CSR) method for the alignment of ontologies. Given a pair of two ontologies, the objective of CSR is to learn patterns of features that provide evidence for the subsumption relation among concepts, and thus, decide whether a pair of concepts from these ontologies is related via a subsumption relation. This is achieved by means of a classification task, using state of the art supervised machine learning methods. The paper describes thoroughly the method, provides experimental results over an extended version of benchmarking series of both artificially created and real world cases, and discusses the potential of the method.