模糊推理的网络舆情观点聚合CA模型研究  

Research on Internet Public Opinion Cluster Model of Cellular Automaton Based on Fuzzy Inference

作　　者：毛乾任[1] 王朝斌[2] 金洪颖[1] 李艳梅[1] 

机构地区：[1]西华师范大学计算机学院,四川南充637009 [2]西华师范大学网络与信息化管理中心,四川南充637009

出　　处：《小型微型计算机系统》2017年第7期1479-1484,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61379036;61300092)资助

摘　　要：为了探究网络舆情的演化机理及传播机制,并考虑到舆情信息以及舆情个体态度的模糊性特点,提出一种基于模糊推理理论的网络舆情观点聚合的CA(Cellular Automaton)模型.定义关注度、态度群落、环境适应强度三种舆情个体属性.将个体属性模糊化,建立模糊推理规则库.把模糊属性反模糊输出的值作为元胞情感状态变化权重,在元胞自动机模型Moore邻域结构下演化并分析个体属性以及邻域环境对网络舆情观点聚合现象的影响.并仿真得到了两种不同模糊定界下的网络舆情观点聚合效果,这些"类聚"状态反映了舆情生命周期的阶段特点,为舆情应对和引导提供参考.In order to explore the evolution mechanism of internet public opinion and its propagation mechanism, this paper puts forward a fuzzy reasoning algorithm of internet public opinion evolution based on the model of CA （ cellular automaton）. Considering the ambiguous characteristics of internet public information and individual attitudes, It proposes some new individual attributes, the concern of Internet public affairs, the community with consistent attitude and the intensity of environmental adaptation. Afterwards, the rules of fuzzy inference are established based on the fuzzy of individual attributes. Under the Moore neighborhood structure of CA model, the evolution results show the influence of individual attributes and neighborhood environment to the evolution of the cluster of network public opinion. With such circumstances of that, the anti fuzzy output value of the fuzzy attribute is used as the weight of the change rule of the cellular state. In the end of this paper,the simulation results show that＂opinion cluster＂ effect comes out among individuals in the development of the internet public opinion. These＂ opinion cluster＂ effects imply the characteristics of corresponding stage of the life cycle of internet public opinion, which provides a reference for guidance and control strategy of the internet opinion.

关 键 词：元胞自动机 网络舆情 模糊推理 观点聚合 MATLAB仿真 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]