Dynamic Features Based Rumor Detection Method

Rumor detection is a hot research issue, and this technology is widely used in various social sites such as Facebook, Twitter and Weibo. The existing rumor detection technologies are mainly divided into two categories: one is traditional machine learning methods such as SVM based on user, content and propagation features, and the other one is neural network model, such as LSTM. In this paper, we propose a novel feature change extraction framework, which combines the advantages of traditional feature detection and neural network detection. We use this framework to extract the change information of dynamic features we define. Compared with relevant studies, our experiment obtains a better identification effect on two microblog datasets. At the same time, our method has better performance on detecting rumors at early stage.