A new model-based Web service clustering algorithm

摘要

Web service has been proposed and widely used as new reusable software function in the Web environment. Web service clustering techniques aggregate Web services with similar features, and can subsequentially improve the qualities of service discovery and recommendation. Existing Web service clustering methods are mainly based on the Web service description information, and ignore the relationship between service, user and provider. To overcome this limitation, this paper proposes a new Web service clustering method, which is based on the heterogeneous service network model. Experimental results based on the WSDream data sets validate that the proposed method has better performance than the existing methods.