A STUDY ON FAKE NEWS DETECTION USING SUPERVISED MACHINE LEARNING CLASSIFICATION ALGORITHMS

Abstract

The Online world has become an essential part of everybody's life in today's society. Almost everyone is using social media to gain information about all around the world. It is an easily accessible source of day-to-day news for almost everyone throughout the globe due to its relatively low cost, ease of access, and rapid dissemination. However, this news comes with a risk of being faulty or fake to mislead the readers. On one hand, it is the most inexpensive, easy, and convenient way of getting information in no time, on the other hand, understanding the background from a headline is vital, the much more difficult task would be predicting the user's purpose; this prediction would be a springboard in the field of natural language processing to identify fake news. False news/information has a tremendous impact on our social lives, in fact, in all fields, particularly politics and education, and organization. The propagation of false information has the potential to create significant social and emotional harm, as well as have potentially dangerous consequences. As a response, automated Fake news detection has become essential for maintaining a sustainable online and social media presence. This study includes the findings of various research being carried out to identify fake news and the possibility of fake content in a particular news article being predicted. In this study, the main motive is carrying out the experiment for the detection of false information by using NLP techniques and five supervised machine learning classification algorithms: Naïve Bayes, Support Vector Machine, Decision Tree, Random Forest, and Logistic Regression and selecting the best algorithm.