O Similarity Measures and their Aggregation in Ontology Matching

The growing number of ontologies on the web has emerged new research issues in the field of semantic web. It is now common to have more than one ontologies on the web for a specific domain. These ontologies may have various types of heterogeneities and therefore it is a challenging task for information retrieval systems to utilize the knowledge of all available ontologies. A number of ontology alignment systems have proposed by researchers in the last decade to bridge this semantic gap. During an alignment process, the entities from different ontologies are compared to find the semantically similar entities. For this purpose, various similarity measures are discussed according to the type of ontological heterogeneities. This paper investigates most common types of heterogeneities exist in ontologies and matchers which can deal with such heterogeneities. Furthermore, the role of aggregation methods in alignment systems have also been analyzed and discussed.