Improved Hybrid Semantic Similarity Algorithm for Terminology Application

With the development of science, people's demand for the technique of query is gradually increased. Especially, the emergency of new terms proposed more demand for query techniques. Therefore, the accuracy of semantic similarity calculation is more important in searching of terms. Now, the hybrid semantic similarity calculation method has been more popular. However, when the expert calculates the semantic similarity, weight values determined are based on expert's experience which has a certain degree of subjectivity and affect the accuracy and objectivity of the semantic similarity calculation. Therefore, this paper proposed an improved hybrid semantic similarity algorithm based on the fuzzy optimization methods. This algorithm could avoid subjectivity for the determined weights and make weights more scientific. In this paper, an example is given for demonstrate how this algorithm can be used for calculating the semantic similarity of volcano terms. Comparing with the old methods, this algorithm can improve query accuracy.