Detecting Malicious Social Bots Through Clickstream Sequence Analysis

This study focuses on interaction kinds, response durations, and navigation patterns in order to identify dangerous social bots using clickstream data analysis. Clickstream sequences are analyzed by an LSTM neural network, which detects anomalies indicative of bot activity. This concept protects platform integrity and improves security by distinguishing between human and bot interactions, particularly in social networks that are susceptible to bot-driven disinformation. Large-scale social networks have verified the model's efficacy, demonstrating its exceptional accuracy in differentiating between legitimate users and malevolent bots.