A semantic relatedness measurement method based on domain ontology considering all relationships

Abstract

The rapid expansion of information and knowledge requires a various applications in computational linguistics and artificial intelligence employing semantic similarity to solve challenging tasks, such as information retrieval, text classification, machine translation, document clustering and so on. Today ontology is an extremely essential approach mainly used to represent acquired knowledge as well as to ensure data and knowledge integration. In this situation, a comprehensive semantic similarity measure is required. We propose a semantic relatedness calculation method that considers all the relationships including common and different attributes relations between concepts. To demonstrate the benefits of exploiting all the relations in domain ontology, we use four method for comparison. The results of our proposed method demonstrate an improvement over the benchmark semantic similarity methods.