Creation of a structured solar cell material dataset and performance prediction using large language models
This study explores the transformative power of big data in materials science, tackling the long-standing issue of data harnessability. The authors introduce a one-step approach that condenses unstructured data from publications into structured formats. By leveraging large language models, this method not only automates the enrichment of existing solar cell datasets but also offers predictive insights into material performance. These advancements underscore the role large language models play in the acquisition of scientific knowledge and the evolution of material science.