Research On Rumors Detection And Rumors Spreading Based On Microblog

In the Web 2.0 era,everyone may express himself on social media.Some messages are rational descriptions of objective facts,while others may be purely fictional fabrications.As one of the largest social media in China,the Sina Blog platform provides users with a great deal of information every day.Many of these messages are rumors that will cause many negative influences.Therefore,finding rumors in time and judging the spread of rumors in advance will not only play an important role in suppressing the spread of rumors,but also have great significance for society and the government.The main work of this paper is divided into two parts:The first part is to detect the specified Blog content and judge its rumors.The second part is based on the content to predict the spread of rumors in the network.In the research work of detecting the rumors attribute of Blog content,this article introduces attention mechanism.Because most of the current text analysis methods cannot completely represent the structured hierarchical information of the text.The ability of attribute analysis to capture context information based on long-term and short-term memory models is limited,and there is a problem of structural information loss.Therefore,it is of practical significance to improve the network structure and apply it to this field.For this reason,this paper proposes a novel hierarchical structure model based on the structure attention mechanism;constructs the HM(ht matrix)structure to store contextual structured information,and introduces the attention mechanism to implement vector selection.At the same time,comparing the model with the original LSTM and ordinary attention model and the existing common methods for proverbial analysis,we found that the structural attention mechanism has a better effect on capturing context-specific information(such as strong negation,etc.)In Blog's rumors spreading research work,this paper introduces game theory,considers the user's selection strategy for specific content of rumors,and quotes the basic game model and game model of responsibilities sharing.From the microblog content point of view,the user is calculated.For the probability of forwarding new rumors,the process of microblogging propagation is analyzed.We combine the game model with the Blog rumor propagation network structure to simulate the spread of rumors.We compared the simulation propagation process with the objective communication situation to verify the effectiveness of the method.In general,this article not only introduces the attention mechanism to detect Blog rumor content,but also establishes a game model based on the process of content simulation to spread the Blog rumors in the real network structure,evaluates the scope of influence of rumors in the network,and verifies the effectiveness of the method.