论文调研了复杂网络的结构特征，基于复杂网络的方法对在线社会网络的多种属性进行了分析，并利用MATLAB仿真工具对在线社会网络数据源进行了仿真，论文重点分析了在线社会网络的度分布、集聚系数、平均最短路径三个基本特征量，仿真实验结果表明在线社会网络具有小世界特性和无标度特性；然后，基于节点收缩的方法对网络节点重要性进行评估，根据节点重要度找出在线社会网络中的核心节点；进而，采用标签传播算法对在线社会网络进行社团划分，改进了算法的结条件，可在不影响划分结果的前提下，大大提高了运算效率。-Studies show that many social networks have the characteristics of the complex network, researching the basic characteristics of the network can help us to understand the various characteristics of social network　By researching the importance of network nodes, we can find the Hub node of the network, it provides the basis of preventing the malicious attacks, protecting critical node or getting a better network transmission　By researching the community structure of network can help us understand the relationship between the individual networks and achieve the maximization of network availability.