Propagating Models Of Cyber Rumor Over Microblog Network And Key Nodes Identification

Cyber rumors are unsubstantiated information spread through social media,as one of the mainstream social platforms in China,microblog not only brings users information sharing,but also facilitates the spread of cyber rumors.With the rapid growth of microblog users,the control of cyber rumors becomes more important.In order to quiet down quickly the microblog rumors,and reduce the propagation range of rumors on microblog network,dynamic models of rumor propagation and methods of key node identification are proposed to control the rumor propagation on microblog network.The main research contents of this thesis are as follows:(1)Aiming at the shortcomings of existing propagation models in user division and transfer relationships,the new UNFR model was proposed.The feature of the model is to introduce a new user state,namely neutral,which takes into account the situation that users will not immediately judge the rumor after contacting it,the model divided user nodes into four categories: unknown,neutral,forwarder and refuter.Node state transitions of model are redefined by considering attenuation effect and forgetting mechanism.Then,the propagation simulation experiments are conducted on microblog network,the experimental results show that increasing the attenuation coefficient and forgetting speed can effectively reduce the spread of rumors,and the effective rumor control strategies are proposed.(2)Considering the situation of microblog users online and offline,the NFR propagating model with online and offline is proposed by combining the nodes division idea of UNFR model,which is closer to the rumor propagation on microblog network.The feature of the model is that it only focuses on the rumor audiences,the model divided users into three categories: neutral,forwarder and refuter,online and offline behaviors of users are also considered.Through the dynamics analysis and simulation experiments of NFR model,results show that restricting the login of some users or closing the accounts of rumor spreading users can effectively inhibit the spread of rumor on the network.(3)The identification of key nodes in the network is studied,the method of neighbor degree averaging and the method of node importance evaluation based on graph convolutional network for identifying key nodes are proposed.Through experimental analysis,the proposed methods have the best comprehensive performance in experimental data sets.Through the identification of key nodes in microblog network and immune experiments,results show that immunizing key nodes can effectively reduce the propagation range of rumors on microblog network.