Zhang, Y., et al. (2025). Automating Materials Science Research: NLP-Driven Pipeline for Data Analysis and Experimentation. In Proceedings of AI4X 2025.
Abstract:
This paper explores how recent advancements in natural language processing (NLP) and generative AI, specifically Text-to-SQL(T2S) and its more general form Text-to-Code(T2C) technologies, can address these challenges, focusing on applications within materials science. We present an integrated pipeline that streamlines material science data retrieval, analysis, and simulation. Specifically, the pipeline utilizes T2S to query structured relational databases (e.g., established repositories like Materials Project, AFLOW, and local repositories) and T2C to generate scripts for extracting data from semi-structured sources (e.g., JSON files, experimental logs). This combination offers a flexible and comprehensive solution for accessing diverse data resources. Furthermore, building on the retrieved data, the pipeline enables automated data analysis with expert-defined functions and supports simulation setup with high-throughput analysis via Text–to–Code.
License type:
Publisher Copyright
Funding Info:
This research / project is supported by the I2R - N.A.
Grant Reference no. : EC-2022-041
This research / project is supported by the National Research Foundation - Campus for Research Excellence and Technological Enterprise (CREATE) programme
Grant Reference no. :