JAMIP: an artificial-intelligence aided data-driven infrastructure for computational materials informatics.

Area of Science:

• Materials Science
• Computational Chemistry
• Data Science

Background:

• Materials informatics accelerates the discovery and design of new materials.
• Machine learning methods applied to large materials datasets are key to this acceleration.
• Specialized infrastructure for data generation, management, and analysis is crucial.

Purpose of the Study:

• To develop an integrated, open-source computational framework for materials informatics.
• To facilitate data-driven materials discovery and design using artificial intelligence.
• To provide a robust platform for high-throughput calculations and machine learning analysis.

Main Methods:

• Development of the Jilin Artificial-intelligence aided Materials-design Integrated Package (JAMIP), an open-source Python framework.
• Integration of modules for materials production, high-throughput first-principles calculations, task management, data handling, and AI-driven data mining.
• Inclusion of an inorganic crystal structure prototype database and machine learning modules for functional materials.

Main Results:

• JAMIP provides an automated, extensible, and reliable infrastructure for computational materials informatics.
• The framework integrates diverse functionalities from data generation to AI-based mining.
• Demonstrated utility in exploring optoelectronic semiconductors, using halide perovskites as a case study.

Conclusions:

• JAMIP is a powerful tool for advancing computational materials informatics.
• The framework supports the acceleration of materials discovery and design through intelligent data utilization.
• Its open-source nature promotes accessibility and further development in the field.