Fake Profile Detection on Social Networking Websites

Abin Varghese, Shyma Kareem

Abstract: The rise of social networking platforms has revolutionized digital interaction but has also led to a surge in fake profiles that threaten user security and trust. These profiles are commonly used for spreading misinformation, phishing, and boosting fraudulent engagement metrics. This paper introduces a machine learning-based solution to detect fake Instagram profiles by analyzing user-centric features such as profile picture presence, bio content, follower-following ratio, and numeric patterns in usernames. Two supervised learning models Random Forest and Decision Tree were trained and evaluated using a dataset of 500 labeled Instagram accounts. The proposed system achieved a detection accuracy of up to 93%, showcasing its potential for scalable, automated fake profile identification. The results highlight machine learning?s effectiveness in improving platform integrity and minimizing human moderation efforts.

Keywords: Fake profile detection, Instagram, Machine learning, Classification, Decision Tree, Random Forest, Profile attributes, Bot detection, social media