A Scalable Approach to Fake News Detection Using TF-IDF and Logistic Regression

Fake news is spreading more widely as a result of the exponential rise of digital news content, which is a major danger to public trust and democratic societies. This project offers a scalable machine learning-based method for identifying false news that uses logistic regression for classification and the Term Frequency-Inverse Document Frequency (TF-IDF) algorithm for feature extraction. We create a binary classification system to differentiate between fake and true news using the Fake Newsnet dataset, which includes sources from PolitiFact and Gossip Cop. Enhancing accessibility and deployment, a Flask-based web interface lets users enter the Uniform Resource Locator (URLs )of news articles and get real-time predictions.