Abstract:
Magnetic materials play a crucial role in numerous aspects of daily life, yet their choice remains limited, and discovering new ones is highly challenging. In recent years, the emergence of machine learning and artificial intelligence has revolutionized materials discovery, offering new hope for identifying novel functional magnetic materials. However, a comprehensive database of magnetic materials is still lacking. In this talk, we address this challenge by leveraging advanced large language models to extract material properties from experimental data reported in peer-reviewed journal articles. The database currently includes more than 30,000 magnetic materials and still keeps growing. Our database is highly inclusive and also encompasses superconductors and thermoelectric materials. We hope this resource accelerates materials discovery and paves the way for a new era in magnetism. No background on magnetism and condensed matter is required for this tak.