Fake News Detection using Bidirectional RNN

Identifying fake news is a growing challenge in the modern era, where misinformation spreads rapidly through online platforms. In this study, an approach which uses Bidirectional Recurrent Neural Networks (RNN) has been taken to detect fake news. The Indian Fake News dataset has been used as a training dataset for the deep learning model. Unlike previous publications, which have focused on a hybrid approach of combining different architectures into a single framework, the BiRNN model was independently used for fake news detection. Additionally, the BiRNN model’s performance was compared with a multilayer perceptron (MLP) algorithm, which yielded suboptimal results. This study focuses on providing insights into the strengths and limitations of recurrent neural networks in applications related to the classification of news.