In this paper, the community clustering problem based on communication data is used to classify the social network, and an effective algorithm for the problem is proposed. A reasonable delivery scheme is given under the premise of satisfying the maximum information coverage and the minimum cost. On this basis, considering the influence of multiple factors, the communication networks given in the question are reclassified.