Time hypergraph neural network rumor detection method and model based on event fusion

The invention provides a time hypergraph neural network rumor detection method and model based on event fusion. The time hypergraph neural network rumor detection method comprises the following steps: constructing a propagation relation graph by using a forwarding user of each tweet; constructing a feature extraction module based on content-social background information by using a convolutional neural network, and obtaining text features based on content and social information; constructing a time-based hypergraph by using propagation structure information of rumors, and obtaining propagation relationship characteristics of the time-based hypergraph by using an event fusion-based graph neural network module; text features and propagation relation features based on content and social information are fused through an information fusion module; and inputting the feature set after information fusion into a rumor classifier for classification prediction. According to the rumor detection method, the information fus