Service Clustering Based on Profile and Process Similarity

The discovery of suitable Web services for a given user requirement is one of the central operations in Service-oriented Architectures. This paper proposes a mechanism to support service discovery via service clustering. Service clustering is aimed at grouping similar services according to the similarity between different services. The procedure of service clustering consists of two phases. The first phase classifies the services into clusters with similar profiles. In order to determine the profile similarity degree, the minimum weights bipartite graph matching is utilized to pair the functionality parameters. The second phase re-classifies the services into clusters with similar process models. Petri net is adopted as a modeling language for the specification of service process model. With the help of Petri net language, the process similarity degree is evaluated via comparing the semantic edit distance. The utilization of service clustering can enable service matchmaker to significantly deploy the discovery of candidate services quickly.