Exploring Application of Large Language Model in Scientific Data Analysis

The text comprehension ability of large model technology can be used for scientific data analysis. By entering the description text of scientific data resources and analysis tools in the prompt words, the large model can perceive the existence of data resources and analysis tools and be able to selectively determine the analysis subjects (i.e. data resources) and analysis tools (i.e. analysis tools) in later analysis tasks. By representing the analysis task as a <background, data, tools, steps> quad and using it as prompt words against the large model, the large model can reuse analysis tools and generate a reliable data analysis workflow by paying attention to the 'steps'.