An Improved Method of Domain Ontology Concept Similarity Calculation based on Semantic Distance

Conceptual similarity calculation has important application value in the fields of natural language processing, information retrieval, knowledge graph and so on. And how to improve the accuracy of conceptual similarity calculation is an important research topic nowadays. Therefore, this paper proposes an improved domain ontology concept similarity calculation method based on semantic distance. This method improves the traditional conceptual similarity calculation method based on semantic distance with three aspects: node distance, semantic coincidence degree, attribute and instance coincidence degree. Experiments on the constructed plant domain ontology were carried out to verify the effect of the proposed method. And the results showed that the maximum semantic similarity value of proposed method is sim (Convolvulaceae, Morning Glory)=0.9285, and the minimum semantic similarity value is sim (Seasonal Flowers, Lilies)=0.3571. The semantic similarity coverage of the proposed method is [0.3571, 0.9285], which was better than the traditional similarity method based on semantic distance. Thus, it is proved that our method has great application potential.