IMPROVING TRUST IN SOCIAL MEDIA PLATFORMS THROUGH ADVANCED MALICIOUS PROFILE DETECTION TECHNIQUES

Abstract

This paper aims to enhance trust in social media platforms by identifying detrimental profiles in a more sophisticated manner through intelligent data analysis and machine learning. The proliferation of social media platforms has increased the likelihood of the occurrence of spam profiles, bots, false accounts, and cybercriminal activities. The veracity of information, the safety of the internet, and the privacy of users are all negatively impacted by these issues. The objective of this investigation is to devise a method for identifying fraudulent profiles by examining user behavior, profile details, posting habits, network connections, and content-related aspects. Machine learning and deep learning techniques, including Random Forest, Support Vector Machine, Decision Tree, and Neural Networks, are employed to facilitate the identification of genuine and fraudulent accounts. The framework employs anomaly detection and feature extraction to identify suspicious activities in real time. When we employed social media datasets for empirical testing, we observed enhanced platform security, fewer false hits, and improved detection performance. This research contributes to the safety of social media platforms by facilitating the identification and prevention of user misconduct.