The Research on Lpa Algorithm and Its Improvement Based on Partial Information

With the growing expansion of data size, in-depth study on the social network clustering algorithm gets prominence. A wide range of researches, from the spectral clustering based on modularity, the hierarchical overlapping clustering algorithm to clustering algorithm based on local information, are mainly based on local neighbor information. The current network node cluster typically has a lower time complexity, thus more suitable for large-scale network data sets. In this paper a detailed analysis of the LPA algorithm based on local information is done. The algorithm is not only able to effectively tap the cluster structure of social networks, but also the concept is more simple and easy to be understood. Further research and analysis show that the LPA algorithm in the network cluster structure of the division can be further enhanced, so this article is committed to this direction. The LPA algorithm is proposed to improve the LPA-the SNA algorithm.