基于局部扩张查询的重叠社区发现

发布时间：2018-05-20 20:51

  本文选题：重叠社区 + 局部扩张　； 参考：《小型微型计算机系统》2015年10期

【摘要】：重叠社区发现源于社交网络、生物神经网络等复杂网络结构分析,在病毒传播防范、网络广告投放和多跳自组路由协议设计等应用中具有重要意义.现有重叠社区发现算法大都是基于静态网络的全局探测,面临复杂度高、灵活性差和健壮性不足等诸多挑战.针对这些挑战,提出一种基于局部扩张查询的重叠社区发现算法—OCLEQ,首先以查询的方式寻找包含特定点的k准团结构,然后基于团结构之间的邻接性实现团的快速扩张,最后定义一个新的度量标准检测和划分遗漏点.仿真实验结果表明,OCLEQ在重叠社区发现的效率和质量上都优于现有方法.
[Abstract]:The discovery of overlapping communities originates from the analysis of complex network structures such as social networks, biological neural networks and so on. It is of great significance in the application of virus transmission prevention, network advertising and multi-hop self-organizing routing protocol design. Most of the existing overlapping community discovery algorithms are based on the global detection of static networks, which face many challenges such as high complexity, poor flexibility and lack of robustness. To solve these challenges, an overlapping community discovery algorithm named -OCLEQ based on locally extended query is proposed. Firstly, the k-quasi structure with specific points is searched by query, and then the fast expansion of clusters is realized based on the adjacency between clusters. Finally, a new metric is defined to detect and divide the missing points. Simulation results show that OCLEQ is superior to existing methods in the efficiency and quality of overlapping community discovery.
【作者单位】： 燕山大学信息科学与工程学院;河北省计算机虚拟技术与系统集成重点实验室;
【基金】：国家自然科学基金项目(61272466,61303233)资助 河北省自然科学基金项目(F2014203062)资助
【分类号】：O157.5;TP301.6
，

本文编号：1916188