Multi-Modal FakeProfile Detection (Text + Image + Network Graph)

Abstract

Social media platforms have become an important part of modern communication and digital interaction. However, the rapid growth of these platforms has also increased the number of fake profiles, spam accounts, bots, and impersonation attacks. Fake profiles are often used for spreading misinformation, online fraud, phishing, cyberbullying, and manipulation of public opinion. Traditional fake profile detection methods mainly rely on single-feature analysis such as text content or account activity, which may not provide accurate detection against advanced fake accounts. This paper proposes a Multi-Modal Fake Profile Detection System using Artificial Intelligence (AI) and Machine Learning techniques. The proposed system combines multiple modalities including textual analysis, profile image analysis, and network graph analysis to identify suspicious profiles more effectively. Natural Language Processing (NLP) techniques are used to analyse usernames, bios, captions, and posts, while deep learning and computer vision methods are applied for profile image verification. Network graph analysis is used to study follower-following relationships and behavioural patterns. The system integrates outputs from all modules using an ensemble fusion model to improve prediction accuracy and reliability. Experimental results demonstrate that the proposed system achieves high accuracy in detecting fake social media profiles. The system also provides explainable outputs and real-time analysis support, making it useful for cybersecurity, social media monitoring, and digital identity verification. Keywords: Fake Profile Detection, Artificial Intelligence, Machine Learning, Deep Learning, NLP, Computer Vision, Social Media Security, Multi-Modal Learning