Research On Community Detection Methods In Service Network

The emergence of service networks provides a new way for service discovery.It can accurately find that services with the same function are the foundation of service discovery,service recommendation and service composition.Therefore,it is an important research direction to improve the division accuracy of service associations on service networks.Existing service community discovery methods mainly use the individual description of the service as the attribute information,and neglect the call and cooperation relationship among the services.In this paper,the service semantics and service relationship between service attributes,service operation invocation relationship,operation and service affiliation are jointly modeled,and the probability generation model of community structure and community semantic closeness is constructed,and then the design and implementation of Service Community Discovery Algorithm Based on Content and Topology.The model consists of two closely connected parts.One is to describe the community structure by using the calling relationship between operations as topological information,and the other is to use the service text as attribute information to describe the community semantics in the service.Nested Expectation Maximization(EM)algorithm is adopted to optimize the parameters and make the model reach the result of global optimization.Experimental results show that the model proposed in this paper is able to accurately accomplish the social network discovery on the service network after completing the data fitting.Compared with other existing clustering algorithms,the proposed model has higher accuracy of community discovery.In summary,the model established in this paper can synthetically consider the two characteristics of service network data with different granularity and generalized community structure,complete the comprehensive modeling of service network structure and service semantic information,and improve the accuracy of community discovery on service network And has good practicality.