Fake News Detector

Abstract

The proliferation of fake news in digital media presents a significant challenge to information integrity. This research explores the application of machine learning, specifically logistic regression, for automated fake news detection using a dataset sourced from Kaggle. Text preprocessing techniques, including tokenization, stemming, and TF-IDF vectorization, were applied to extract features from news articles. A logistic regression model was trained on the processed data to classify articles as real or fake. The model achieved high accuracy rates of 98.68% on the training set and 97.67% on the testing set. Additionally, a user-friendly Streamlit web application was developed for real-time prediction of fake news. This study demonstrates the efficacy of logistic regression in combatting misinformation and contributes to enhancing information credibility in digital media. Top of Form Key Words — Fake news detection, machine learning, logistic regression, TF-IDF vectorization, text preprocessing, information integrity.