Fake accounts detection system based on bidirectional gated recurrent unit neural network

<span>Online social networks have become the most widely used medium to interact with friends and family, share news important events or publish daily activities. However, this growing popularity has made a target for suspicious exploitation such as spreading of misleading malicious information, making them less reliable trustworthy. In paper, fake account detection system based on bidirectional gated recurrent unit (BiGRU) model is proposed. The focus been content users’ tweets classify twitter user profile legitimate fake. Tweets are gathered in single file transformed into vector space using GloVe word embedding technique order preserve semantic syntax context. Compared baseline models long short-term memory (LSTM) convolutional neural (CNN), results promising confirm that BiGRU classifier outperforms 99.44% accuracy 99.25% precision. To prove efficiency our approach obtained were compared Word2vec under same conditions. Results performs best Twitter accounts only feature.</span>