A new Structural Similarity Measure for Ontology Alignment

several methods have been suggested for Ontology Alignment problem, each exploits some measures and uses of their combinations in order to come up with higher quality of alignment. A group of such measures are especially based on structural position of the entities in ontologies, the corresponding measures of which are named Structural Measures. This paper is about to present a new method for computing Structural Similarity, based on the notion of Information Content. The method employs various aspects of the structure of ontologies to recognize related entities. We evaluate the proposed measure using standard methods of Precision and Recall, as well as a new one based on Sensitivity Analysis from Data Mining. Evaluation results show superiority of proposed structural measure over the rest of the methods taking part in the presented tests.