Research On Rumor Source Estimation Of Online Social Network Based On Exoneration And Prominence

It is in the era of rapid development of information technology.Various online social networks are developing vigorously,accelerating the speed of information diffusion,and expanding the scope of information impact.On the one hand,it provides convenience for people,and on the other hand,it also provides a breeding ground for rumors and other malicious information.Rumors are likely to have a negative impact on society and cause social turbulent.Therefore,it is of great significance to research the origin of the rumors and determine the key information for the spread of rumors and to curb the development of rumors under different conditions.Based on the related theories of complex networks and the law of rumors spreading,this paper studies the tracing of rumors.The main contents are as follows:1)Based on the heterogeneous SI propagation model,considering that the source of the rumor is the earliest infected node in the network,that is,the node with the oldest node age,comprehensively the relationship between the node's exoneration and age,and proposes a source traceability algorithm based on exoneration and prominence.At the same time,in order to reduce the computational cost,select high betweenness centrality nodes are regarded as suspicious sets.Finally,simulation experiments conducted in several synthetic and real networks show that the algorithm can quickly and effectively identify the source of rumors.The average error distance of the traceability results in multiple networks is less than 1 hop,which is compared with other traceability algorithms in the experiment.Certain superiority.At the same time,in a highly sparse network,the algorithm performs well.2)Researched the problem of dual-source traceability in heterogeneous networks.Based on the structural characteristics of the complex network community,on the basis of modularity,using optimized spectral analysis methods,the dual-source infection network is divided into two communities,and single-source tracing work is carried out independently in each community.Finally,in several real networks,comparing the performance of different traceability algorithms,the results show that the average error distance of the algorithm is the smallest.In the Karate network,the correct detection probability is as high as 0.595,which is compared with other rumor traceability algorithms in the experiment,the performance is better.3)The problem of multi-source tracing under the unknown situation of rumor spreading model is studied.Considering the fact that the source of real network rumors is not unique,multi-source traceability is completed based on the tag value information after multiple iterations and the continuous source identification algorithm.Finally,do simulation comparison experiments in several synthetic and real networks and average error distance and correct detection probability are its evaluation criteria.Results show that the performance of the proposed algorithm is better and more stable than other algorithms when the number of rumor sources increases.