Community Mining in Signed Networks Based on Dynamic Mechanism

摘要: Topology structure of the networks is crucial to understand their structural and functional characteristics. Networks may be social networks, internet, political that involve natural language connections, so on. Topics community detection in signed have attracted a lot attention recent years. In this paper, novel network model based on dynamic mechanism proposed detect networks. Similarities between nodes are defined select out positive neighbors with higher similarity negative lower similarity. Initial state values randomly generated. With evolution time, would cluster together get away. Finally, all into different groups stable values. By Lyapunov stability theory, proved uniformly stable. Lots real synthetic tested verify our method. Besides, thorough comparisons demonstrate presented method superior three state-of-the-art algorithms.