Social media multi-modal rumor detection method based on propagation heterogeneous graph modeling

Abstract

The invention discloses a social media multi-modal rumor detection method based on propagation heterogeneous graph modeling, and the method comprises the steps: extracting text and image information through a pre-training model at a feature extraction stage, and capturing the structural information of social media through a graph convolution neural network model based on deep learning; according to the method, the information can be allowed to be propagated through the constructed graph network according to the propagation characteristics of the social media, so that richer information is obtained, limited marking data and a large amount of unmarked data can be fully utilized, and resource waste caused by manual marking is reduced. And in the rumor detection stage, using a softmax classifier to perform rumor detection by using the features after the network structure information and the multi-modal information are fused. Through the method provided by the invention, rumor detection canbe automatically, quickly and accurately realized, so that spreading of false information and non-real speech and adverse effects caused by spreading of false information and non-real speech are reduced.

Description

technical field [0001] The invention relates to the technical field of cyberspace security, in particular to a social media multimodal rumor detection method based on propagation heterogeneous graph modeling. Background technique [0002] With the development of society, traditional social media has become an important source for users to share information, and social media has an influence that cannot be ignored in information dissemination. But what followed was the viral spread of all kinds of false information. The spread of rumors caused public panic, disrupted social order, affected public opinion, and manipulated the focus of the public, becoming a great social instability factor. Therefore, proposing an effective method for automatically detecting false rumor information is of great significance for maintaining social stability and cyberspace security. [0003] In order to suppress the flood of rumor information in social media, the academic community has proposed r...

AI Technical Summary

Problems solved by technology

However, none of the current detection schemes takes these factors into consideration, so the detection accuracy needs to be improved

Images