Rumor Situational Awareness Based On Fractal Theory

The development of network and information technology has spawned a variety of social media platforms that people can share and access information very quickly and easily.The network is like a highway that is well-connected.It greatly facilitates the exchange of information,but it breeds all kinds of unexpected problems.The network rumors ragged with them.The spread of rumors will disrupt public order,cause panic to the majority of netizens,and bring severe challenges to network governance.China is the country with the largest number of Internet media applications.Once the Internet rumors spread,it will have extremely bad impact on all Internet users in China and pose a serious threat to the long-term stability of our society.Therefore,it is imperative to effectively reduce the spread of rumors,improve user relations,and promote social harmony.It is especially important to further identify and predict the development of rumors by studying the different characteristics of rumors and non-verbal communication.Based on this,firstly,we obtain the propagation situation map of the rumor,and based on the nonlinear characteristics of the propagation situation,use the multifractal principle to establish a rumor situation recognition model based on multifractal.The model collects a certain type of rumor and non-rumor propagation situation map on the network,uses the box dimension calculation method of multifractal dimension to calculate the multifractal dimension of the two types of models,and uses the multifractal dimension as the eigenvector to train the SVM.A classification model that uses this model to identify rumor information.Secondly,we select the propagation situation map of a certain rumor,calculate the propagation situation by R/S analysis method,and obtain the Hurst index.According to the range of Hurst index,it is determined whether the rumor situation has the characteristics of a fractal distribution.The fractal interpolation algorithm is used to establish the rumor situation prediction model,and the situation of the next stage is predicted based on the trend of the previous period.Through model construction and experimental tests,the results show that the multifractal rumor situation recognition model can quickly and effectively identify rumorsfrom the propagation situation,and the accuracy rate can reach 0.85.At the same time,the fractal interpolation prediction model can fit the rumor propagation situation and predict the propagation situation of a period,and verify the feasibility of the fractal interpolation prediction model.