Fake News Classification from Online Social Network Comments using BILSTM Deep learning Model

Social networks (SNs) are becoming more essential news providers. Online users' connections propagate information faster than conventional news channels like newspapers and TV. However, they also facilitate the propagation of fake information news, notably via social media. Politicians and media have utilized "fake news" often in recent years. Large volumes of material that the public can't trust or that may easily confuse and mislead them might engender public mistrust in media. This article examines if SN users' network structure might help Twitter detect bogus news. This project employs a knowledge base of tweet patterns and numerous algorithms to determine emotion and authenticity. In addition, the data was lemmatized and tokenized, which improved news article authenticity. BILSTM deep learning algorithm is proposed to address this fake news detection from tweets. Some baseline classifiers were applied to the data and compared.