Automated multi-model fake news classifier

The widespread increase in fake news, whether created by humans or machines, has a negative impact on society and individuals on both a political and social level. The rapid rotation of news in the age of social media makes it difficult to assess its authenticity quickly. As a result, automated fake news identification tools have become a necessity. To solve the aforementioned problem, a hybrid Neural Network architecture is used, which incorporates the capabilities of CNN and LSTM, as well as two separate dimensionality reduction methods, PCA and Chi-Square. We'll use data from the Fake News Challenges (FNC) website, which includes four different forms of stances: agree, disagree, discuss, and unrelated. The aim of this study is to figure out what a news article's body is in relation to its headline using different deep learning and ML models.