The Method for Semantic Similarity Based on Concept Distance
Xinying Chen, Guanyu Li, Heng Chen, Yunhao Sun, Wei Jiang · 2019
Semantic matching is an important problem of service discovery.In order to find effective services, a method for semantic concept similarity is proposed.The method calculates the concept similarity between the parameter concepts of services by directly using the distance relationship between the concept nodes in the classification tree.And uses the nonlinear function to calculate the similarity and redefine the concept-based similarity between concepts.The new method effectively solves problems in existing algorithms and further improves precision.Finally, theoretical analysis and experimental result reveals the validity of the proposed method.