USING MACHINE LEARNING IN DETECTING FAKE NEWS

In a world that has been greatly affected by the Coronavirus pandemic and more recently by the armed conflict between Russia and Ukraine, the flow of information is constantly increasing and at the same time the veracity of this information raises a big concern, and this makes the topic of fake news a problem of major interest. Our paper proposes a tool for fake news detection using different models of machine learning developed over a Fake News Corpus. Neural networks have proven to be the most effective method, reaching an accuracy of over 90%, but also Naive Bayes can be an excellent solution for classifying text data. Besides these two, we also developed and analyzed other models based on Naive Bayes and k-Nearest Neighbors. The results are promising and show that the problem of fake news can be managed by machine learning algorithms.