This project delivers a complete, end-to-end solution for detecting fake news by leveraging a powerful machine learning pipeline. Using the substantial WELFake dataset, which contains over 72,000 news articles, the project first applies rigorous natural language processing techniques, including tokenization, stemming, and lemmatization to transform raw text into clean, analyzable data.

Several models were evaluated to find the most effective classifier for this task. The project tested a high-performance XGBoost model, a transparent Decision Tree, and a baseline Naive Bayes classifier. The models' performances were measured by their accuracy in distinguishing real from fake news, with XGBoost achieving the highest accuracy at 97%, followed by the Decision Tree at a respectable 93%, and Naive Bayes at 88%.

Given its superior performance, the XGBoost model was selected for the final application. The project's true strength lies in its practical implementation: an interactive prediction tool. This feature allows anyone to input a news article and receive an instant classification. Crucially, the tool integrates LIME (Local Interpretable Model-agnostic Explanations) to showcase why a decision was made by highlighting the specific words that most influenced the prediction. This emphasis on explainability demystifies the AI, turning a "black-box" model into a transparent and trustworthy tool for navigating the complex landscape of modern media.