Abnormal Behavior Analysis Based on Truth Discovery and Machine Learning
Qiang Huang, Yun He, Bin Wu, Hanqi Zhao, Weihua Lian · 2022 Global Conference on Robotics, Artificial Intelligence and Information Technology (GCRAIT) · 2022
Abnormal behavior analysis has been extensively studied in social network security. Many abnormal users often post advertising links, spread false news, and even launch malicious attacks. Existing solutions of abnormal behaviors analysis mainly focus on low-dimensional features. In this paper, we propose an advanced analysis scheme for abnormal behaviors, providing the following properties: behavioral, semantic and temporal based feature extraction, multi-dimensional feature learning based on LightGBM, BiLSTM and SVM, and abnormal behavior analysis based on truth discovery. The performance evaluation shows that the proposed scheme can achieve 96.7% accuracy and is efficient and feasible.