启发式动态社区挖掘算法研究与实现

针对社会网络的动态特征,应用多模态函数优化和粒子群优化算法的基本思想,引入社区种子和社区主题的概念,分层进行社区的挖掘.首先对复杂网络中存在的固定联系进行社区挖掘,构建基本社区结构;然后分析社区内容,根据社区内节点之间的隐性行为特征定义社区主题,精分细化社区结构直到结构稳定.实验证明,该算法极大地提高了社区挖掘的精度,降低了运算复杂度.并且该算法能够有效地保持社会网络中社区的多样性,加速社区内节点收敛,快速寻找到稳定的社区结构.

Research and implementation of heuristic dynamic community discovery algorithm

In terms of the dynamic features in social network,the concepts of community seed and community theme based on the instruction of multimodal function optimization and particle swarm optimization algorithms are introduced to hierarchically discover the social community.Firstly,the immovable connections are used to dig community structures in complex networks and construct the fundamental community structures.Secondly,the community context is analyzed and the community theme is defined according to the recessive behavioral characteristics of nodes in the same community.Then,the community members are refined until the structure is steady.Experimental results show that this algorithm can effectively preserve the diversity of communities in social network,accelerate the convergence of nodes in communities and quickly find the steady social structures.