摘要
社区结构作为真实复杂网络所普遍具有的一个重要的拓扑特性,最近10年内得到了广泛而深入的研究。为解决社区挖掘策略时间复杂度过高、缺少与用户交互等问题,讨论了社交网络节点中心度、度的幂律分布等特性,提出了"关键子网络"和"社区框架"的概念,设计了社区框架挖掘算法MCF(Mine the Community Framework)和社区框架钻取算法DCF(Drill Down the Community Framework),其中MCF算法用于挖掘社交网络的社区框架,DCF用于对社区框架进行钻取,从不同粒度展现社区结构。实验结果和实验分析表明,MCF算法能够在较短时间内挖掘出反映复杂网络社区状态的社区框架,DCF算法可以以用户交互方式实现高质量的社区划分。
As an important topological characteristic which the real complex networks commonly have,community structure has been widely and thoroughly studied in recent 10 years. To solve the problems of community mining strategy that its time complexity is too high and lacks the interaction with users,etc.,we discussed the node centrality,node's power-law degree distribution and other characteristics of social networks,and proposed the concepts of ' critical sub-network' and ' community framework'. Moreover,we designed the community framework mining( CFM) algorithm and the community framework drilling( CFD) algorithm. Among them,the CFM algorithm is used to mine the community framework of social networks,and CFD is used for drilling the community framework and to demonstrate the community structure from different granularities. Experimental results and analysis showed that,in a relatively short time the CFM algorithm could be used to mine out the community framework reflecting the complex networks community state,while the CFD algorithm could implement highquality community partition in the way of user interaction.
出处
《计算机应用与软件》
CSCD
2016年第7期83-87,共5页
Computer Applications and Software
基金
国家自然科学基金项目(61170052)
关键词
社交网络
社区结构
节点中心度
社区框架
社区质量
Social networks
Community structure
Node centrality
Community framework
Community quality
