Comparison of Supervised Learning Algorithms for Fake News Detection

Today, national and international information and news have greatly increased in volume. Almost all people have access to everything they want through devices and the internet. Information is just a click away. But besides this very good aspect, there are also negative sides. A negative aspect, but also the most important, is the fact that news can be distorted by those who process it and post it online. Thus, you can read information, but it may not be the real one, thus becoming misinformed. Various researches have been carried out that have tried to discover methods and algorithms to detect this fake news, but not all algorithms used are effective in detecting whether online content is false or true. There are machine learning algorithms that can detect such news and tell whether the information is true or false. This paper aims to discover which supervised learning algorithm is the most efficient and best at detecting fake news from real ones. Thus, several supervised learning algorithms were trained on a dataset taken from Kaggle and finally the most efficient algorithm that can be used for fake news detection will be presented.