Large language model accurately predicts online chat derailments

Tsukuba, Japan—Online chat rooms and social networking platforms frequently experience harmful behavior as discussions drift from their intended topics toward personal conflict. Traditional predictive models typically depend on platform-specific data, limiting their applicability and increasing implementation costs.

In this study, the researchers applied a zero-shot prediction method to LLMs to detect conversational derailments. The performance of various untrained LLMs was compared to that of a deep learning model trained on curated datasets. The results showed that untrained LLMs achieved comparable, and in some cases superior, accuracy.

These findings suggest that platform operators can implement effective moderation tools at reduced cost by leveraging general-purpose LLMs, supporting healthier online communities across diverse platforms.

_{Original Paper}

Title of original paper:
Zero-Shot Prediction of Conversational Derailment With Large Language Models

Journal:
IEEE Access

DOI:
10.1109/ACCESS.2025.3554548

_{Correspondence}

Associate Professor YOSHIDA, Mitsuio
Institute of Human Sciences, University of Tsukuba

NONAKA, Kenya
Doctoral Program in Risk and Resilience Engineering, Degree Programs in Systems and Information Engineering, University of Tsukuba

_{Related Link}

Institute of Business Sciences

Master's / Doctoral Program in Risk and Resilience Engineering