Abstract: This paper studies the algorithm for detecting community structure of complex network based on clustering, analyzes the similarity measure method between vertices. It proposes Mapping Vertex into Vector(MVV) algorithm, which converts all vertices in network into vectors. It converts the nodes into the data structure suitable for clustering algorithms. It compares the different clustering algorithms and similarity measure method, the results show that MVV algorithm can improve the ability of detecting complex networks community.

Key words: complex network, community structure, clustering, data mining

摘要： 对基于聚类技术的复杂网络社团发现算法进行研究，分析网络中结点间的相似性度量方法，提出把复杂网络中的结点转化为向量的顶点到向量映射(MVV)算法，把网络中的结点转化成适合聚类算法的数据结构形式。对不同聚类算法及相似性度量方法的性能进行比较分析，结果表明，MVV算法可以提高发现复杂网络中社团的能力。

关键词: 复杂网络, 社团结构, 聚类, 数据挖掘

CLC Number: 

• TP311