Abstract

We present MinMod, one of the largest knowledge graphs of historical mining data. MinMod is built using scalable machine learning methods to extract hundreds of thousands of records from mining reports, databases, and tabular data in articles, and to normalize and integrate the results into a unified knowledge graph. It also provides tools that enable end-users to explore, curate, and leverage the data to support predictions of new sources of critical minerals. In this demo, we walk through the process of extracting and integrating data into MinMod, demonstrate data exploration and curation, and showcase tools such as the Grade & Tonnage model that assist scientists in mineral assessments.

Date

March 7, 2026

Authors

Craig A Knoblock, Binh Vu, Basel Shbita, Pothula Punith Krishna, Namrata Sharma