Twitter Bot Detection

Twitter bot detection aims to identify automated accounts spreading misinformation and manipulating online discourse. Current research emphasizes leveraging both the textual content of tweets and the network structure of user interactions, employing machine learning models like graph neural networks (GNNs) and language models (LMs), often combined or used in a knowledge distillation approach. These advancements improve accuracy and efficiency compared to older methods, addressing challenges like class imbalance and the evolving tactics of sophisticated bots. The resulting improvements in bot detection have significant implications for social media platforms, researchers studying online influence, and efforts to combat the spread of disinformation.