Rumor Suppression Strategy Considering Community Structure In Online Social Networks

In recent years,with the rapid development of Internet technology,various online social networking platforms make it very convenient for people to obtain information.Therefore,people’s use,trust and dependence on the Internet have gradually increased,and the impact of the Internet has penetrated into all corners of people’s lives.However,because there is no specific requirement for publishing information in social networks,users have full right to speak,and it is also very easy to forward and spread the information they see,while social network platforms cannot accurately identify all false information,so online social networks have also caused widespread spread of online rumors.At the same time,the huge scale of users of online social networks,as well as their virtual,free,interactive and other characteristics,make the spread of online rumors more rapid and extensive,which has a great impact on people’s lives and cyberspace security.Therefore,designing effective suppression strategies to curb rumor diffusion has become an important research direction in the field of cyberspace governance.The existing research mainly suppresses rumors from two ways:spreading positive information to resist rumors;Block rumors by blocking a certain number of users or links between users.This paper aims to curb the spread of rumors in online social networks by blocking links between users.First of all.this paper establishes an IC Model Considering Community Structure(CSIC)of online social networks.The model assumes that the spread of rumors is not only affected by the influence of nodes,but also depends on the attribute similarity between nodes,thus more truly describing the spread of rumors in social networks.On this basis,this paper proposes three rumor suppression algorithms based on link blocking,that is,to curb rumor diffusion by blocking links between users.They are respectively Path separation blocking algorithm for global networks(PSG).Path merging blocking algorithm for global networks(PMG)Path merging blocking algorithm in community(PMC).For this reason,this paper introduces the concept of the maximum influence tree,so that each node’s influence on rumor propagation is limited to its local influence area of the maximum influence tree,and the size of this local influence area can be adjusted by changing the regional influence threshold of the node,so as to balance the running time and inhibition effect of the algorithm.Among them,PSG algorithm starts from the whole network and blocks each path with the greatest impact,so that it can search the links in the network that will cause the maximum spread of rumors in a short time to block them;PMG algorithm starts from the whole network and blocks the maximum influence path of each node.Compared with PSG algorithm,PMG algorithm has better blocking effect,but its running time is slightly longer;PMC algorithm determines the popularity of rumor information in the community according to the similarity between rumor information and community users’ attributes,so as to block rumors for specific communities.In this paper,the above algorithms are simulated in three real social networks.The experimental results show that the running time of PSG algorithm and PMG algorithm decreases with the increase of node area influence threshold,but the effectiveness of the algorithm decreases slightly;In large-scale social networks,the blocking effect of PMG algorithm is better than that of PSG algorithm;The performance of the algorithm based on seed node is also better than that based on non seed node.Finally,PMC algorithm can quickly and effectively identify and block the communities interested in rumors with obvious attribute characteristics.The research in this paper provides some decision-making basis for relevant departments and online social network platforms to formulate effective rumor blocking strategies.