RUMOR SITUATION RECOGNITION BASED ON MULTIFRACTALS

Abstract

In order to effectively identify rumors and further block their propagation, this present study established a rumor situation identification model based on multifractals and support vector machine (SVM), which considered the fractal theory and the situation of rumor propagation. First, the representative rumors and nonrumors about food safety were selected. The multifractal analysis was used to calculate the generalized fractal dimension spectrum of the two target states. After this, the generalized fractal dimension values were used as the input target eigenvector for the SVM, which was subsequently trained to obtain the best classification results. The experimental results show that the model can quickly and effectively identify rumors from the communication situation, thus providing further help for governance.