基于EASD风险准则的社交网络信息传播模型  

Social Network Information Dissemination Model Based on EASD Risk Criteria

作　　者：徐建民 袁国强[2,3] 王苗 XU Jian-min;YU AN Guo-qiang;WANG Miao(School of Cyber Security and Computer,Hebei University,Baoding 071002,China;School of Management,Hebei University,Baoding 071002,China;School of Big Data Science.Hebei Finance University,Baoding 071051,China)

机构地区：[1]河北大学网络空间安全与计算机学院,河北保定071002 [2]河北大学管理学院,河北保定071002 [3]河北金融学院大数据科学学院,河北保定071051

出　　处：《模糊系统与数学》2020年第4期108-121,共14页Fuzzy Systems and Mathematics

基　　金：河北省社会科学基金资助项目(HB18TQ007)。

摘　　要：本文通过可信性分布函数刻画信息传播中的风险性因素,并建立三种不同的社交网络信息传播风险模型。将可信性理论和绝对半偏差风险评价准则应用到社交网络信息传播问题中,进而考虑从信息传播流量期望值的两侧分别定义不同的风险评价指标来分析信息传播中的部分风险性因素。从数值算例的结果可以看出三种信息传播风险模型的最优值随着关键参数的整体变化波动比较小,最大的相对误差没有超过1.002%,这也充分说明在关键参数变化时,模型的最优值是比较稳定的。三种信息传播风险模型均考虑的是无向全连通社交网络问题,但对于部分有向非连通的实际社交网络,模型的实用性效果就会相对比较差。新模型不仅丰富了社交网络分析的研究方法,而且将可信性理论引入到社交网络信息传播风险问题中,提高了部分主观不确定性因素的处理效率。The risk factors in information dissemination are characterized by the credibility distribution function in this paper,and three different risk models for information dissemination on social networks are established. The credibility theory and the absolute semi-deviation risk evaluation criteria are applied to the information dissemination problem of social networks, and then different risk evaluation indicators are defined from the two sides of expected value for information dissemination flow to analyze some risk factors in information dissemination. From the results of numerical examples, it can be seen that the optimal values of the three information dissemination risk models fluctuate less with the overall change of key parameters, and the maximum relative error does not exceed 1.002%. This also fully shows that the optimal values of the models are relatively stable when the key parameters change. Three kinds of information dissemination risk models are undirected fully connected social network problems, but for some of the directed non-connected social networks, the practical effect of the models will be relatively poor. The new models not only enrich the research methods of social network analysis, but also introduce credibility theory into risk problem of information dissemination on social networks, and improve the efficiency of dealing with some subjective uncertainties.

关 键 词：信息传播 可信性理论 期望绝对半偏差 等价形式 Lingo软件 

分 类 号：O159[理学—数学]