Considering the optimization of such indexes as convergence and coupling degree in service identification algorithm, we proposed a service identification algorithm based on Weighed Network Optimized Clustering (WNOC) by using the relational degree model and the method of network topology clustering, introducing the adjacency parameter of clustering, and taking the coupling-convergence degree as the optimized objective function. Simulation was made with Matlab to a certain example, and the optimal clustering result was selected according to the coupling-convergence degree optimized model. The result proves the validity of the method in service identification.