Local optimization overlapping community discovery algorithm combining attribute features

When the data scale is too large and the social network is incomplete, the traditional local optimization overlapping community discovery algorithms mostly rely too much on the structural characteristics of the social network and ignore the community node attributes. For this reason, in this paper, we propose a locally optimized overlapping community discovery algorithm that integrates attribute features. We first obtain several community core nodes by combining node attributes and structure information to improve the density peak fast search algorithm. After that, we use the structural fitness between core nodes and non-core nodes, community homogeneity, and community heterogeneity values to establish the objective function, and gradually discover the community where the core node is located according to the minimization of the objective function difference in the iterative process. Finally, we merge communities with a high degree of overlap to reduce redundant communities while improving the quality of community division. Experimental results show that the proposed algorithm can significantly improve the quality of locally optimized overlapping community discovery.