Fake News Detection using Deep Recurrent Neural Networks

The widely spread of fake news has significantly impacted our life in politics and economics. To solve this problem, different researchers have proposed various machine learning and deep learning models. However, most of them detect fake news without desired accuracy. Therefore, we proposed a deep learning framework that classifies fake news from real ones with 99.82% accuracy. This BiLSTM model was trained and tested on a fact-checking dataset. Furthermore, we used different model evaluation metrics like precision, recall, Fl-meassure, execution time to prove the efficiency of our model.