Study of Bot detection on Sina-Weibo based on machine learning

Abstract

Accompany with the growth of Sina-Weibo users, mendacious Bot users also emerge, which lead to network environment pollution and lower management efficiency. This paper focuses on Sina-Weibo users, extracts the effective features of Bot user through behavior analysis and features study. Then based these features, Bot user identification model is trained by machine learning process and model performance evaluation. The result shows that these extracted features have satisfactory discrimination and the identification model has good performance.