A Hybrid Algorithm for Solving Two-Part Division Problem in Network Community Detection

Network representation is a convenient and intuitive abstraction for analyzing the massive interacting data. Some topological characteristics of the network have been found in the past decade, and community structure is the typical one of them. Community detection has become a hot topic in complex network analysis. In the paper, a hybrid algorithm is presented for solving such problem. At first, we take the max-degree node as the "core" node. Then, the diffusing operation is per-formed on it to produce two preliminary partitions. Subsequently, an EO-based adjustment step is used for generating the final partitioning with high quality. In addition, four real-world networks are used for validating the efficiency and effectiveness of our hybrid algorithm. The experimental results show that the proposed hybrid algorithm is a promising solution for solving the community detection problem both in precision and efficiency.