Research On The Propagation Model Of Rumor & Anti-Rumor Game For Multi-Message

With the emergence of online social networks,the speed of information dissemination has been greatly improved.The emergence of social networks has brought many benefits to people,but the emergence of rumors has caused huge hidden dangers to the society.Rumor refers to false statements or information published without official sources and without being reviewed by relevant staff.Rumors are false in nature,and any sensitive rumor topic may cause social unrest.Social networks are the main way for online rumors to spread,and users can release all kinds of information without review.Nowadays,rumors have become the fuse of public opinion,and the spread of rumors has greatly affected the social order.Therefore,how to intelligently predict the spread trend of rumors and explore the key factors that affect users’ forwarding is crucial for public opinion departments to timely discover and clarify rumors.In this thesis,rumors in multiple messages are taken as a starting point to predict and analyze the propagation trend of rumors in multiple messages from the perspective of user group behavior and network state.The core work of this thesis is as follows:1.At the level of user behavior.In view of the fact that there are multiple messages in the communication space of a rumor topic,and the current research lacks consideration of the influence between multiple messages,and considering the complex symbiotic and antagonistic relationship of multiple messages in the communication process,this thesis proposes a rumor-refuting communication model for multi-message game.Firstly,the representation learning algorithm is used to express the user’s interest characteristics and network structure characteristics quantitatively.At the same time,considering the compound iterative cascade of multi-information rumor,this thesis proposes a rumor-refuting game driving mechanism of multi-message compound game from the perspective of multi-message.Finally,the Graph Convolutional Network(GCN)is used to convolution the non-Euclidean structure data of social Network.This paper proposes a multi-message Game rumor propagation model MMGAME-GCN(Multi Message Evolutionary Game Theory GCN)based on graph convolution neural network,which can well predict users’ forwarding behavior.2.At the level of network state.Aiming at the complex influence among multiple messages,this thesis measures the message influence according to the multiple linear regression algorithm,quantifies the complex influence among multiple messages effectively,and excavates the factors affecting the rumor spreading from multiple angles.Aiming at the symbiosis and antagonism between messages under multiple messages,the multi-message game driving mechanism is constructed based on evolutionary game theory,so as to better explore the competition and cooperation between multiple messages.Based on the theory of infectious disease model,this thesis introduces a new population state — multi-message state M,and comprehensively considers the polymorphism of rumor information in multi-message.Finally,this thesis proposes a multi-message oriented rumor transmission dynamics SIMR model,which can more accurately perceive the situation of online rumor transmission.In addition,the validity of the model is verified by using real data set published by Sina Weibo.Experimental results show that the proposed method can not only predict users’ forwarding behavior and rumor propagation trend in multi-message network,but also reflect the cooperation and competition between multiple messages.