Fake News Detection Using Machine Learning Technique

The rapid expansion of digital news platforms has increased the spread of fake news, posing serious challenges to information credibility and public trust. The purpose of this work is to develop and evaluate an efficient machine learning-based approach for fake news detection. A comparative analysis of four classification algorithms—Logistic Regression, Decision Tree, Gradient Boosting, and Random Forest—is conducted using a dataset of 44,878 news articles comprising both fake and genuine content. After preprocessing the news text with normalization, URL, punctuation, and special character removal, TF-IDF vectorization is used to convert it into numerical characteristics. Accuracy, precision, recall, and $\mathbf{F 1}$-score are used to assess the model's performance. The Decision Tree classifier achieved the best accuracy of 99.53 %, according to the data, which show that all models reach accuracy above 98 %. Our result shows that machine learning methods can be used in automated content verification systems to reliably identify fake news when paired with efficient text preprocessing.