Adithi Pai S

Abstract The rapid growth of online media has accelerated the circulation of misinformation, posing serious risks to public trust, social stability, and democratic processes. Detecting fake news remains a challenging task, particularly because traditional machine learning approaches such as Support Vector Machines and Naïve Bayes rely on shallow linguistic cues and often fail to recognize sarcasm, implicit tone, or evolving writing styles.