Early Detection and Prevention of Malicious User Behaviour on Twitter Using Deep Learning Techniques

Abstract

Malicious user behavior on Twitter, such as spam, harassment, and disinformation, can degrade the platform's integrity and user experience. Traditional detection techniques often struggle to detect emerging threats in real time. This paper presents a deep learning-based approach for the early detection and prevention of malicious user behavior on Twitter. By analyzing user interactions, tweet content, and engagement patterns, the system predicts and identifies potentially harmful users before they escalate their behavior. The model leverages Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTM) networks, incorporating features such as linguistic characteristics, temporal patterns, and user metadata. Experimental results show that the proposed system achieves a 92.3% accuracy, outperforming traditional methods in both detection rate and prevention capability. This system can assist in early intervention, significantly reducing the harmful impact of malicious users on the platform.