TWEETSENTINEL: MACHINE LEARNING FRAMEWORK FOR MALICIOUS BOT IDENTIFICATION

Abstract

Social media platforms, particularly Twitter, have become prime targets for malicious bots that spread misinformation, spam, and manipulate public opinion. Detecting these bots is essential for maintaining platform integrity, user trust, and information reliability. This study introduces TweetSentinel, a machine learning-based framework designed to accurately identify malicious Twitter accounts. The system analyzes user behavior patterns, tweet content, temporal activity, and network interactions, employing supervised and ensemble learning techniques to distinguish between human users and automated bots. Experimental results demonstrate that TweetSentinel achieves high accuracy, precision, and recall, outperforming traditional detection methods. The framework provides a scalable, adaptive solution for real-time bot detection, contributing to safer and more trustworthy social media environments.