复杂网络演化中的“熵减点”研究:以微博传播网络的演化为例  

"Entropy Reduction Point" in Complex Networks:Taking the evolution of micro-blog transmission network as an example

作　　者：沈乾[1] 黄远[1] 马宁[1,2] 刘怡君[1,2] 

机构地区：[1]中国科学院科技政策与管理科学研究所,北京100190 [2]中国科学院自然与社会交叉科学研究中心,北京100190

出　　处：《数学的实践与认识》2015年第3期282-290,共9页Mathematics in Practice and Theory

基　　金：国家自然科学基金委重大研究计划"非常规突发事件应急管理研究"培育项目:"非常规突发事件中社会舆论形成;演化;引导;干预的系统建模与仿真分析"(91024010;91324009);中国科学院创新团队项目(KACX-YW-1011;GH13041);中国科学院科技政策与管理科学研究所青年基金(Y201201Z06)

摘　　要：基于标准网络结构熵这一度量"有序性"的指标,以微博传播网络为例,首次将这一宏观指标运用到网络的微观演化研究中.首先,根据"网络大V"参与情况的不同,将微博传播网络分为"无V型网络"、"单V型网络"和"多V型网络".其次,依托实际数据,在标准结构熵的视角下,分别运用枚举法、跟踪法、介入法探讨了这三种网络的演化特征,发现了网络中不同节点的微观演化对宏观"有序性"的差异性影响,其中"大V"节点具备明显的"熵减"效应.最后,归纳提出"熵减点"的相关概念,并提出一套用来识别复杂网络重要节点的"熵减点判别法".方法相较于传统方法具有一些优点且适用性较强,可以扩展到其他复杂系统与复杂网络研究中,为网络结构熵理论的实际应用提供了一些思路.Based on the standard network structure entropy as a measurement index of ordering, taking the micro-blog transmission network evolution as an example, for the first time to apply this macro indicators in the study of the network＇s micro evolution. Firstly, according to the participation of the ＂big V＂ microblog users, the micro-blog transmission networks could be divided into three evolutionary status including ＂zero V network＂, ＂single V network＂ and ＂mutual V network＂. Secondly, the enumeration method, the tracking method and the intervention method are respectively adapted to the analysis of three kind of networks. It is found that the ＂big V＂ points have an entropy reduction effect. Finally, the definition of ＂entropy reduction point＂ and a method of entropy reduction point identification in the complex network are promoted. Compared to traditional methods, this method has some advantages. And, The method is suitable for other complex network when the definition of structure entropy is adapted. The method is helpful for the empirical study of structure entropy theory.

关 键 词：标准结构熵 微博传播网络 熵减点 网络演化 

分 类 号：O157.5[理学—数学]