MatSKRAFT: A framework for large-scale materials knowledge extraction from scientific tables
Scientific progress increasingly depends on synthesizing knowledge across vast literature, yet most experimental data remains trapped in semi-structured formats that resist systematic extraction and analysis. Here, we present MatSKRAFT, a computational framework that automatically extracts and integrates materials science knowledge from tabular data at unprecedented scale. Our approach transforms tables into graph-based representations processed by constraint-driven GNNs that encode scientific principles directly into model architecture. MatSKRAFT significantly outperforms contemporary frontier large language models, achieving F1 scores of 89.33% for property extraction and 71.35% for composition extraction, while processing data 6-496 × faster—compared to the fastest and the slowest models, respectively—with modest hardware requirements. Applied to 66,267 tables from more than 45,500 research publications, we construct a comprehensive database containing 509,281 entries, including 104,000 compositions and a broad range of material properties that expand coverage beyond major existing databases. This systematic approach reveals previously overlooked materials with distinct property combinations and enables data-driven discovery of composition-property relationships forming the cornerstone of materials and scientific discovery.
Authors
- Mohd Faiz Mohammad Zaki (ORCID: https://orcid.org/0000-0002-4551-3470)
- N. M. Anoop Krishnan (ORCID: https://orcid.org/0000-0003-1500-4947)
- Kausik Hira (ORCID: https://orcid.org/0009-0003-5071-7667)
- Mausam
Institutions
- Indian Institute of Technology Delhi (IN)
Publication Details
- Journal
- npj Computational Materials
- Published
- 2026-11-16
- DOI
- https://doi.org/10.1038/s41524-026-02359-5
- Citations
- 5
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 3.50
Funders
- Microsoft
- Alexander von Humboldt-Stiftung
- Intel Labs
- Indian Space Research Organisation
- Ministry of Education, India
- Indian Institute of Technology Delhi
- Board of Research in Nuclear Sciences
- Wipro
- Google Research
- IBM Research