Credibility of Social-Media Content Using Bidirectional Long Short-Term Memory-Recurrent Neural Networks

Fake news is false information that is created, circulated and endorsed similarly as real news. Fake news and the dissemination of fake stories is nothing new to us; it has existed even before the internet was invented. But now due to social media, any person sitting at any corner of the world can create and spread fake news or rumors within minutes, which has the potential to create serious negative implications for society As a consequence, the identification of false news has arisen as a modern study field that is gaining more and more interest every day. In this article, we'll concentrate on the legitimacy of social network news. With the aid of Long ShortTerm Memory (LSTM)-recurrent neural networks, we expect to present a false news identification model. Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) are two deep neural networks that have been shown to recognizedifferent dynamic trends in textual results. The Dataset used is Fake news Detection Dataset which is publicly available by kaggle, which contains social media news articles that we have used in training our model. As compared to other deep learningstrategies such as Convolutional Neural Networks (CNNs), Gated Recurrent Units (GRU), Unidirectional Long Short- Term Memory - Recurrent neural networks, and other Bi- directional Long Short-Term Memory models, we find that ourproposed Bi-directional LSTM model outperforms in terms of accuracy and performance.