NFDI4Culture Metadata API

How can data from different disciplines be compared, integrated, and meaningfully combined? To further explore this question, representatives from the Mathematical Research Data Initiative (MaRDI), National Research Data Infrastructure for Materials Science & Engineering (NFDI-MatWerk), and NFDI4Culture came together once again for an interdisciplinary workshop. Building on the first meeting (see short report), the focus this time was on identifying points of intersection and shared approaches: what happens when mathematics, materials science, and the humanities look at the same object from different perspectives?

Since the previous workshop, each consortium has developed its methodological perspective further. A shared vision has become clear: within the NFDI, all disciplines meet through a common language – the language of data. This shared foundation is enabling new research questions and innovative approaches to understanding historical materials.

Medieval stained glass – particularly objects from the German Corpus Vitrearum Medii Aevi (CVMA) – emerged as a promising use case. Representatives from the three consortia presented their respective approaches, research questions, and attempts to answer art-historical inquiries about stained glass through data-driven methods.

A key goal of the workshop was the development of a joint, sustainable data corpus. Discussions included methods for capturing window geometries using polygon outlines, analysing glass fragments via high-resolution image data ("TIFF maps"), and identifying material properties – such as chemical composition, microstructure, or optical behaviour – as potential "fingerprints" to help determine provenance or workshop origin.

The workshop demonstrated impressively how the combination of cultural-historical expertise, data-based mathematical analysis, and material science methods can create innovative, interdisciplinary ways of acquiring knowledge, which were met with great enthusiasm. The results of the workshop will soon be published in a scientific article.