Safeguarding Twitter from Bots using Temporal Analysis

The increasing presence of automated bots on social media platforms, particularly Twitter, poses challenges such as misinformation, spam, and malicious activities. While traditional machine learning algorithms remain useful, they often fail to detect sophisticated bots that mimic human behavior. This work focuses on developing a model for Twitter bot detection using temporal analysis of user activity data. The proposed approach emphasizes improving detection accuracy and robustness. The model’s performance is evaluated against traditional machine learning methods as well as ensemble models to demonstrate its effectiveness in identifying bots. Also, the performance of each model based on key metrics like accuracy, precision, recall and F1-Score are recorded.