Detecting Bot-Generated Text by Characterizing Linguistic Accommodation in Human-Bot Interactions
This addresses the problem of detecting malicious bot-generated text (e.g., for misinformation or hate speech) for online platforms and security applications, representing an incremental improvement over existing detection methods.
The paper tackles bot-generated text detection by showing that analyzing how people respond to bots (rather than the bot text itself) yields more robust detection across datasets and models, while also providing insights into linguistic differences between human-human and human-bot conversations.
Language generation models' democratization benefits many domains, from answering health-related questions to enhancing education by providing AI-driven tutoring services. However, language generation models' democratization also makes it easier to generate human-like text at-scale for nefarious activities, from spreading misinformation to targeting specific groups with hate speech. Thus, it is essential to understand how people interact with bots and develop methods to detect bot-generated text. This paper shows that bot-generated text detection methods are more robust across datasets and models if we use information about how people respond to it rather than using the bot's text directly. We also analyze linguistic alignment, providing insight into differences between human-human and human-bot conversations.