Dresden 2026 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
MM: Fachverband Metall- und Materialphysik
MM 22: Data-driven Materials Science: Big Data and Workflows III
MM 22.3: Talk
Wednesday, March 11, 2026, 10:45–11:00, SCH/A216
Building a FAIR Community around Parsing — •Nathan Daelman1, Alvin N. Ladines1, Esma Boydas1, Martin Kuban1, Bernadette Mohr1, Sascha Klawohn1, Rubel Mozumber1, Christina Ertural2, Silvana Botti3, Joseph F. Rudzinski1, Lauri Himanen1, and FAIRmat Team1 — 1Inst. für Physik, Humboldt-Universität zu Berlin — 2Department of Materials Chemistry, Federal Institute for Materials Research and Testing, Berlin — 3RC-FEMS and Faculty of Physics, Ruhr University Bochum
NOMAD [nomad-lab.eu][1, 2] is an open-source data infrastructure for materials science data. One of its most praised features is how NOMAD allows for direct ingestion of various software output formats. This gives data producers access with minimal effort to the whole toolkit infrastructure system regardless of their choice of simulation code. As the NOMAD community extends into related scientific disciplines, parsing procedures should grow alongside and empower casual users to contribute too. To this end, I will be presenting two new parsing frameworks: (i) Mapping Annotation which connects code-specific formats to the NOMAD interoperable schema, while gracefully handling syntatic concerns; (ii) an agentic LLM interface for hooking up third-party parsers via the Model Context Protocol (MCP). Finally, I will highlight how both approaches fit into NOMAD Plugins and NOMAD Actions.
[1] Scheidgen, M. et al., JOSS 8, 5388 (2023).
[2] Scheffler, M. et al., Nature 604, 635-642 (2022).
Keywords: FAIR data; research data management; parsing; LLM; ab initio