基于最大可达概率的虚假信息溯源问题研究

Misinformation source estimation based on the maximum reachable probability

随着互联网的持续普及,虚假信息传播的速度和趋势在增加,广泛的传播将会不断影响并损害国家、社会及个人的利益,负面影响会变得难以控制.因此,本文基于最大可达概率研究了在社交网络中的三阶段虚假信息溯源问题,对于无向且连通的感染图G(V,E,p)而言,给出了最大可达概率的定义,从而进一步得到网络中每个节点的接近可达概率和最长可达路径值.除此之外,本文基于IC模型获取源候选节点集(IC-OCNS),再基于最大可达概率对网络中源候选节点集进一步筛选,提出三阶段可达概率算法(TSRA)来更好地解决社交网络中虚假信息的溯源问题.最后,在实验阶段,本文在现实世界的三个社交网络数据集中进行实验比较,验证算法的可行性和误差率,该研究可以为网络安全中虚假信息的控制和应对提供管理和决策支持.

With the continuous popularization of the internet,the speed and trend of dissemination of misinformation are increasing.Widespread dissemination will continue to affect and harm the interests of the country,society and individuals,and negative effects will become difficult to control.Therefore,based on the maximum reachable probability,this paper studies the three-stage misinformation source estimation problem in social networks.For undirected and connected infection graphs G(V,E,p),the definition of the maximum reachable probability is given to further obtain the near reachable probability and the longest reachable path value of each node in the network.In addition,this paper obtains a candidate source node set based on the IC model(IC-OCNS),and then further screens this set in the network according to the maximum reachable probability.A three-stage reachability algorithm(TSRA)is proposed to better solve the problem of misinformation source estimation in social networks.Finally,at the experimental stage,this paper conducts experimental comparisons on three social network data sets in the real world to verify the feasibility and caculate the error rate of the algorithm.This research can provide management and decision support for the control and response of misinformation in network security.