Analysis of Machine Learning Algorithms via Detection of Fake News

In the modern political climate, fake news is a growing and legitimate threat to our institutions and all voters. Fake news articles are those which are “intentionally and verifiably false”. This project is aimed at implementing combinations of various feature extraction techniques along with various Machine Learning algorithms from distinct categories and for the purpose of detecting fake news articles through their content. The results of this supervised binary text- classification problem will be compared and ranked. Kaggle which is owned by Google LLC and is a community of data scientists and machine learning engineers which will fulfill our requirement of a reliable source that provides us with a verifiable dataset of real and fake news. To mirror the real-world environment, the quantity of fake news articles in the dataset, will be substantially less than the amount of real news articles. A data set with approximately 85: 15 ratio will be used.