Metric of intrinsic information content for measuring semantic similarity in an ontology

  • Authors:
  • Md. Hanif Seddiqui;Masaki Aono

  • Affiliations:
  • Toyohashi University of Technology, Hibarigaoka, Tempaku, Toyohashi, Japan;Toyohashi University of Technology, Hibarigaoka, Tempaku, Toyohashi, Japan

  • Venue:
  • APCCM '10 Proceedings of the Seventh Asia-Pacific Conference on Conceptual Modelling - Volume 110
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Measuring information content (IC) from the intrinsic information of an ontology is an important however a formidable task. IC is useful for further measurement of the semantic similarity. Although the state-of-art metrics measure IC, they deal with external knowledge base or intrinsic hyponymy relations only. A current complex form of ontology conceptualizes a class (also often called as a concept) explicitly with the help of the hyponymy classes and the asserted relations and restrictions. Therefore, we propose a modified metric for measuring IC intrinsically taking both the concept-to-concept and the concept-to-property relations. We evaluate our system theoretically and with experimental data. Our evaluation shows the effectiveness of our modified metric for extracting intrinsic information content to measure semantic similarity among concepts in an ontology.