基于谱优化社区划分的双信源溯源算法  

Identifying Dual Information Source with Community Partition Based on Spectral Optimization

作　　者：廖艺 王友国[1] 朱亮[2] LIAO Yi;WANG You-guo;ZHU Liang(School of Science,Nanjing University of Posts and Telecommunications,Nanjing 210046,China;School of Telecommunications&Information Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210003,China)

机构地区：[1]南京邮电大学理学院,江苏南京210046 [2]南京邮电大学通信与信息工程学院,江苏南京210003

出　　处：《计算机技术与发展》2020年第12期72-76,82,共6页Computer Technology and Development

基　　金：国家自然科学基金资助项目(61771256)。

摘　　要：在线社交网络的飞速发展给人们带来便捷服务的同时也给谣言的肆意传播提供有利的平台,若不加以制止,将会严重扰乱社会秩序。因此,如何快速准确地识别谣言源具有重要的实际意义。考虑到社交网络的社区结构特性,即社区内节点连接紧密,社区间节点连接松散,通过分析扩散快照和网络拓扑结构,提出一种结合社区划分的谣言溯源算法。在模块度的社区划分算法的基础上,基于优化的谱分析方法将感染图划分成两个社区,然后运用谣言中心性的溯源算法在两个社区内分别进行单信源溯源,将双信源溯源问题近似分解为两个独立的单一信源溯源问题。为验证该算法的有效性和准确性,对比不同的网络拓扑结构和不同的中心性估计量,仿真实验结果表明该算法能够快速有效地识别谣言源。The rapid development of online social networks not only brings convenient services to people,but also provides a favorable platform for the rampant spread of rumors.If the rumor spreads wantonly,it will seriously disrupt social order.Therefore,how to quickly and accurately identify the rumor source has important practical significance.Considering the characteristics of the community structure of social networks,that is,the nodes within the community are closely connected,while the nodes between the communities are loosely connected.Through analyzing infection snapshot and network topology,a rumor identifying algorithm based on community partition is proposed.On the basic of the modular community division algorithm,the infection graph is divided into two communities based on the optimized spectral analysis method,and then the rumor centrality identifying algorithm is used to identify single source in the two communities respectively.The problem of identifying dual information source is approximately decomposed into two independent single source traceability problems.To verify the effectiveness and accuracy of the algorithm,comparing different network topologies and different centrality estimators,simulation results show that the proposed algorithm can quickly and effectively identify the source of rumors.

关 键 词：社交网络 社区划分 谱优化 溯源 谣言中心性 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]