Express Analysis of the Structural and Electronic Properties of Nanomaterials Using Big Data, Large Language Models and Generative Artificial Intelligence

The possibility of using large language models, generative machine learning and big data methods for predictive analysis of the electronic properties of nanostructures based on semiconductor materials is considered, without limiting the generality of this approach for other crystalline materials. The conceptual solution in the form of a cross-platform software application with an advanced search generation capability for applying this approach to working with scientific data arrays is described. Preliminary test results showed a positive result of using the developed solution to predict semiconductor properties (bandgap width, Fermi energy) of a reference material. Prospects for the development and implementation of big data methods, large language models, and generative artificial intelligence in the context of modern trends in materials science are discussed.