Uncovering the community structure in complex networks is one of the most interesting issues in the social analysis. There are many good effect algorithms in complex networks. Inspired by those famous ideas, a multi-paths relationship matrix is defined and Newman’s modularity is modified to “strength modularity”. Then we develop an algorithm based on strength modularity to extract the optimal community structure. Experiments for several real-world social networks were valid which confirmed the feasibility and effectiveness of the algorithm based on strength modularity.