Published Paper Details:

The authenticity of information has become a long standing issue affecting businesses and society in social media. On social networks, the reach and effects of information spread occur at such a fast rate and spreading false information on social media may cause real world impacts. Rumors on social media may cause public panic and negative impact on individuals. So it is necessary to makes the automatic detection of rumors and blocking the user who continuously spreads rumor. This system detect rumor by learning user representation(User behaviour) by graph convolutional networks(GCN) and learning content semantics and propagation clues of rumored tweets. GCN efficiently capture node features and graph structure features in graph structure data and it helps to learn the user information. The proposed system uses user based features, content based features, propagation based features to detect rumors. The main objectives of this system is develop a social network without any fake news spreading. It taken twitter based social network to predict the fake news spreading system. In this system fake news detection is done by using Recurrent Neura Network and Graph Convolutional Network. This model experiments on two real-word datasets show that this method is more superior to the state-of-the-art on rumor detection.