复杂网络中社团发现算法的研究

摘要

基于复杂网络模型,将数据挖掘中的聚类分析方法应用到社团发现中,提出了结合模块度的基于层次聚类的社团发现算法.由层次树得到的社团结构层次清晰,仿真实验证明,利用该算法,当信号传播次数取值为3时社团划分准确度最高.%Based on complex network,clustering analysis method in data mining is applied to the research of community detection.A new measured method for node similarity--node dissimilarity coefficient in multi-subnet composited complex network is proposed.A community detection algorithm in complex network based on hierarchical clustering is proposed.By using this algorithm,community classification derive from the hierarchical tree is very clear.Experiments prove that when the number of signal propagation is 3,high accuracy rate of community classification are received in network.