Research On Rumor Spreading Model And Feature Fusion Detection Method_?

With the rapid development of science and technology,the Internet has entered thousands of households,social networks have gradually become an indispensable part of people's lives.In the Web2.0 era,everyone can publish their own opinions and ideas through social media.As a well-known online social platform in China,Sina Weibo has a massive registered user base,and the amount of information disseminated on the platform every day is huge.In this information,spam and even rumors will inevitably be generated,and the spread of this information will have a negative impact on the network environment and even society.Therefore,timely detection of rumors and effective measures to suppress the spread of rumors are of great significance to maintaining network stability and social stability.This article mainly conducts research from two parts:the first part is to analyze the spread of rumors through modeling;the second part is to detect the content of Weibo to achieve the purpose of rumor discrimination.The main work is as follows:(1)Based on the SIR infectious disease model,the S_uS_eITRtransmission model is proposed to analyze the spread of rumors.This model divides susceptible persons into two states according to whether they are affected by media propaganda and adds real information disseminators,that is,the state of rumor defamers,and analyzes the influence of various parameters in the propagation process on the stability of the system through dynamic equations and MATLAB simulation experiments.And put forward the strategy that the government or related institutions can increase the intensity of rumor rejection and strengthen ideological education to effectively suppress the spread of rumors.(2)This paper proposes a new micro blog rumor detection model with attention mechanism.Considering that the existing CNN models ignore the influence of features on the output results in the research of rumor detection,attention mechanism is added to CNN model to give different weights to the extracted features according to their influence on the output results,so as to improve the efficiency of rumor detection.The experimental results show that the accuracy rate of the model with attention mechanism is 96.8%,and it also has good performance in accuracy rate and recall rate.(3)A rumor detection method based on multi feature and multi model fusion is proposed.Considering that the selection of rumor features and models in the research of rumor detection is mostly single,a feature model fusion method is proposed.Compared with single feature,multi feature extraction has more advantages in detection,and model fusion can make the defective models complement each other,so as to improve the detection accuracy.Experiments show that the method of multi feature and multi model fusion can improve the accuracy of rumor detection.To sum up,this paper not only proposes a new propagation model for rumor propagation analysis,but also proposes a new rumor detection model,hoping that this research can help the existing rumor research work.