A Method Measuring Semantic Relatedness for Concepts in Ontology Base
L. Wang
2006-01-01
Abstract:The notion of semantic relatedness between concepts in Ontology finds use in many contexts, for example, in search engines, web service semantic matching, and clustering. Many methods relating semantic relatedness or similarity have been suggested that are classified to four sorts: counting words/concepts similarity based on WordNet, obtaining semantic distance only by their structure such as graph methods, getting semantic similarity by machine learning from large quantity of individuals and making probability statistic such as informatics method, some special fields such as using lattices for semantic matching and so on. In this paper, we introduce an easy-to-use, flexible algorithm measuring semantic relatedness between concepts in OWL, based on ontology structure and concepts’ semantic relation, which does not need other lexical base such as WordNet, nor machine learning from quite large quantity of individuals. It just occurs in a domain ontology base. Firstly we analysis the concepts’ relationship in OWL and classify them corresponding to DL (Description Logic) into subsume, non-subsume, equivalence, reference. Because in DL, subsume is an important relation, the detailed subsuming theories and computing lemma are given. According to these principles, we propose our method measuring classes’ semantic relatedness sufficiently using ontology hierarchical structure and semantic relation. A prototype of our approach is currently under development as part of the CSAOR (Content&Situation Awareness Ontology Knowledge Routing) project. At last a case is given and the comparison of our measures against some measures is given showing our method’s validity.