一种面向动态网络的社团检测与演化分析方法

针对制约动态网络演化分析方法发展的社团演变模式挖掘问题，设计了基于指向性变异策略和变邻域搜索算法的静态社团检测算法与基于匹配度和社团生存周期的社团演化分析算法，并采用在时刻上运行静态社团检测算法、在时序上运行社团演化分析算法的策略，提出了一种面向动态网络的社团检测与演化分析方法。并用Zachary空手道俱乐部网络和Power网络验证了该方法的可行性和有效性。

New Community Detection and Evolution Analysis Method for Dynamic Networks

Aiming at the mining problem for evolutionary community patterns, which restricts the development of evolution analysis methods for dynamic network, this paper designs a static community detection algorithm based on a kind of directed mutation strategy and variable neighborhood search algorithm, and a community evolution analysis algorithm based on compatibility and community {\rm{lifetime}}. Through adopting a strategy that runs static community detection algorithm on the moment and community evolution analysis algorithm on the sequential, a new community detection and evolution analysis method for dynamic network is proposed. In the experiment, the feasibility and effectiveness of the proposed method are verified by Zachary karate club network and power network.