一种基于注意力机制与双向GRU的谣言检测方案

With the popularity of online social networking, network rumours have a very bad impact on the society. How to accurately identify network rumors on online social platforms is particularly important for maintaining social order. This paper proposes a network rumor detection model that combines the attention mechanism and Bi-GRU, and proposes a new method to process the comment data of the original topic, so as to combine the original topic and the comment data to construct better features. Finally, experiments show that the model\'s judgment of network rumors has improved a lot in indicators such as accuracy and recall.