DefinePK

Abstract

Social media platform proliferation of fake accounts, for example, on Instagram, is a major problem regarding user security, user engagement, and credibility of the platform. Malicious activities like spamming, fraud, and impersonation are commonly done through fake accounts. This paper discusses different machine learning methods for identifying fake accounts on Instagram. The research entails data preprocessing methods, feature extraction techniques, and a comparative analysis of the various machine learning models utilized for classification. The performance of these models on real-world datasets is analyzed, including the challenges and limitations of the current solutions. The results indicate that the Support Vector Machines (SVM) and Random Forest (RF) achieve the best performance in terms of accuracy and F1-score. The paper finishes by considering the future directions of enhancing fake account detection with the integration of deep learning and real-time detection systems.
Keywords : Instagram, Fake Account Detection, Machine Learning, Data Preprocessing, Classification Models, Social Me- dia Security.