Fake News Detection Using Machine Learning Techniques

One of the most important initiatives has been the widespread adoption of internet use. People use the Internet for a wide range of reasons. These people are well-versed in the use of numerous social media channels. These web platforms allow any user to publish content or disseminate news. Neither the individuals nor their content is authenticated on these networks. Individuals occasionally utilise these mediums to disseminate deceptive information. The exponential spread of fabricated information has emerged as a crucial concern in the contemporary culture that heavily relies on information. The spread of deceptive and inaccurate information presents significant concerns to public perception, democratic procedures, and the overall welfare of society. In response, researchers have progressively relied on machine learning (ML) algorithms to create automated systems for identifying fraudulent news. Motivated by this, we propose to create a robust and efficient system that can accurately detect and counteract the spread of misinformation. In this study, we employed eight different ML models to detect fake news, and their performance is assessed on two real-world datasets obtained from Kaggle. For feature extraction, these models utilize the Term Frequency-Inverted Document Frequency (TF-IDF) technique. Among the other eight models, the Passive Aggressive Classifier (PAC) demonstrated the highest performance, achieving an average accuracy of 97.26%.