Fake Social Media Profile Detection
Parth Bagul∗, Piyush Jadhav†, Dipali Bari‡, Pooja Malpure§, Mr. Pramod Gosavi¶
∗†‡§UG Students, ¶Associate Professor,
Department of Computer Engineering, SSBT’s College of Engineering and Technology, Jalgaon, Maharashtra, India
Abstract—In this paper, we present a system designed to detect and flag fake social media profiles through a combina- tion of machine learning techniques, behavioral analysis, and natural language processing. Dubbed ”FakeProfile,” the system leverages state-of-the-art technologies such as scikit-learn for model training, spaCy for linguistic feature extraction, and a neural network-based classifier for robust prediction accuracy. The system analyzes a variety of features, including posting patterns, profile metadata, friend/follower ratios, linguistic cues, and image inconsistencies, to distinguish between authentic and inauthentic user accounts.
Aimed at enhancing platform integrity and user safety, FakeProfileX provides real-time analysis and threat scoring, allowing platforms to take proactive measures against bot- driven or malicious activity. It supports manual verification and automated workflows, giving administrators the flexibility to review flagged profiles or implement auto-removal protocols. A built-in dashboard—developed using Python, Streamlit, and Matplotlib—visualizes profile behavior over time and highlights anomalies in user activity.
Designed with scalability and adaptability in mind, Fake- ProfileX is suitable for integration into existing social media ecosystems, offering a modular and extensible architecture that can evolve alongside emerging forms of online deception. By combining behavioral intelligence with machine learning, the system represents a meaningful step toward safeguarding digital communities and ensuring more authentic online interactions.
Index Terms—Generative Artificial Intelligence, Gesture Recognition, Human Computer Interaction, Computer Vision.
