A Machine Learning Framework for Real-Time Rumor Detection on Social Media Platform

This study develops a machine learning-based system for the real-time detection of rumors on social media, particularly focusing on Twitter. We evaluated several models including Logistic Regression, Support Vector Machine (SVM), Naive Bayes, Random Forest, Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and an Ensemble model combining Random Forest, SVM, and Logistic Regression. The CNN model demonstrated superior performance with the highest accuracy of 88%, closely followed by the Ensemble and SVM models at 87%. Naive Bayes reported the lowest accuracy at 82%, indicating challenges with complex social media text. The effectiveness of the CNN is attributed to its advanced capability to capture complex linguistic and contextual nuances, significantly enhancing classification accuracy. Furthermore, the system includes a user-friendly interface developed with Streamlit, facilitating swift and accurate information verification by users. The robustness of these machine learning models, particularly CNN, marks a significant advancement over traditional rumor detection methods, offering a scalable and effective solution for combating misinformation on social media platforms.