[1J4-OS-10b] OS-10

[1J4-OS-10b-05] Automatic Summarization of Multiple Comments with ChatGPT

〇Tatsuki Ito¹, Wataru Sunayama², Shun Hattori² (1. Graduate School of Engineering, The University of Shiga Prefecture, 2. School of Engineering, The University of Shiga Prefecture)

Keywords:ChatGPT, LLM, Automatic Information Summarization, Information Abstraction

In recent years, the opportunities for data analysis have increased, and the integration of diverse information requires an enormous amount of time and effort. In addition, the emergence of large language models has made it possible to automate sophisticated tasks that were once thought to be difficult to automate. However, in automatic opinion aggregation, the traditional method of opinion extraction is the mainstream, and automatic abstraction of opinions is rarely seen. In this study, we propose a system that automatically aggregates multiple comments by abstracting them using ChatGPT. We aim to create an environment in which users can easily acquire generalized knowledge from multiple comments. Through preliminary experiments, we compared the automatically generated summaries by the system with those by humans, and evaluated the quality of the summaries.