Rumour detection technology based on the BiGRU_capsule network

The widespread use of social media enables people to quickly and conveniently obtain and share information, and it also brings about the rapid spread of false information. A large number of online rumors flooded the social network media, which had a serious impact on people’s lives and social stability. The research on the detection technology of network rumors has become a hot research topic in recent years. The capsule network can clearly view the hierarchical structure of the data by viewing the activated path. This paper proposes a rumor detection model based on a hierarchical capsule network. The model can use static routing and dynamic routing to obtain clear input signals and accurately determine data information; the model fully takes into account the spatial structure characteristics of the text, effectively identify the content of the text. Therefore, this model can effectively improve the accuracy of rumor detection. This experiment uses twitter public datasets, under different conditions of dynamic routing and static routing, compared with the existing typical rumor detection models, the accuracy of the proposed model is improved by 6.1% and 5.6%.

Li Dong

Papers

Twitter Rumor Detection Technology Based on Hierarchical Capsule Network