PRE-PROCESSING ONTOLOGIES TO IMPROVE THE RESULTS OF MATCHERS

Abstract– Ontology matching is one of the most important topics in many fields of semantic web
research. The focus of ontology matching is on discovering semantic correspondence between
entities of differently modelled ontologies of the same domain. Although there have been many
efforts on developing ontology matching tools, the results of matching process is not appealing. In
this paper, a new approach is proposed in order to get a better result from matching process. In this
approach a pre-processing phase is added to matchers for analysing and modifying input
ontologies. Refactoring rules are utilized on detected inappropriate patterns. Matching algorithms
are then applied on the refactored ontologies. To evaluate our proposed approach two well-known
matchers, RiMOM, and ASMOV, are used. Evaluation results show that our proposed approach
improved the quality of the matching process with respect to standard evaluation measurements
such as precision, recall, and F-measure.