Abstract
In this study, we developed a method for generating systematic and reproducible research data based on the knowledge learned by ChatGPT. Specifically, we designed and introduced a prompt that extracts the required data from ChatGPT to visualise the complex adversarial and cooperative relationships among major military powers. Using this prompt, we obtained structured data and applied multi-dimensional scaling to visualise the positional relationships between countries. The results showed that average values from multiple responses enhance the reproducibility of the data.
This study contributed to further analyses by transforming ChatGPT’s responses into usable research data.
https://ror.org/03jqhsn85 
