A review on Fake Profile Detection Methods Using Machine Learning Approaches

Abstract

The increasing use of social media and digital platforms has led to a surge in fake profiles, which are often created for malicious purposes such as spreading misinformation, committing fraud, or manipulating public opinion. These fake accounts pose a serious threat to user safety, trust, and platform integrity. Traditional rule-based methods and manual moderation are no longer sufficient to handle the scale and sophistication of these deceptive profiles. As a result, researchers have turned to machine learning (ML) techniques to develop more effective and automated detection systems. This review explores various machine learning approaches used in fake profile detection, including supervised learning models like Decision Trees, Random Forests, and Support Vector Machines, as well as unsupervised techniques such as clustering and anomaly detection. It also highlights the application of deep learning and graph-based models for analyzing complex user behavior and network patterns. Key stages such as feature extraction, model training, and evaluation are discussed, along with the importance of selecting relevant features such as posting frequency, sentiment of messages, and social network metrics. The review also addresses current challenges such as class imbalance, adversarial tactics, and dataset limitations. In conclusion, machine learning offers a powerful and scalable solution for detecting fake profiles, but continuous innovation is needed to stay ahead of evolving threats.