Automatic lexical annotation applied to the SCARLET ontology matcher

Abstract

This paper proposes lexical annotation as an effective method to solve the ambiguity problems that affect ontology matchers. Lexical annotation associates to each ontology element a set of meanings belonging to a semantic resource. Performing lexical annotation on the ontologies involved in the matching process allows to detect false positive mappings and to enrich matching results by adding new mappings (i.e. lexical relationships between elements on the basis of the semantic relationships holding among meanings). The paper will go through the explanation of how to apply lexical annotation on the results obtained by a matcher. In particular, the paper shows an application on the SCARLET matcher. We adopt an experimental approach on two test cases, where SCARLET was previously tested, to investigate the potential of lexical annotation. Experiments yielded promising results, showing that lexical annotation improves the precision of the matcher. © 2010 Springer-Verlag Berlin Heidelberg.