Fake Profile Detection Using Machine Learning

Abstract: This paper proposes an With the rapid growth of social networking platforms, fake profiles have become a major concern as they are often used for spreading misinformation, scams, phishing, and other malicious activities. Detecting these fake profiles is essential to ensure user safety and platform integrity. This paper presents a machine learning–based approach to identify fake profiles by analysing behavioural, structural, and content-based features. The system collects user data such as friend count, posting frequency, profile completeness, message patterns, and interaction rates. These features are then preprocessed and fed into supervised learning algorithms such as logistic regression to classify profiles as genuine or fake. The model’s performance is evaluated using metrics like accuracy, precision, recall, and F1-score. Experimental results demonstrate that machine learning techniques can effectively distinguish fake profiles with high accuracy, thereby providing a reliable solution for social media security and trust enhancement.

This project aims to design and implement an effective fake profile detection system using various machine learning algorithms. The system utilises a Twitter profile dataset containing both genuine and fake user accounts. Genuine accounts are categorised as TFP and E13, while fake ones are classified as INT, TWT, and FSF. Several attributes— such as the number of followers, friend count, frequency of status updates, account creation date, and profile completeness—are extracted and analysed to distinguish between real and fake accounts.