Brief Report | 19. December 2024

Connecting Research Data from History and Musicology: A collaborative NFDI4Memory and NFDI4Culture Workshop and Hackathon

By Alexandra Büttner, M. A. , Linnaea Söhn, M. A. and Dr. Heike Fliegl

The Culture Knowledge Graph not only enables the exploration of research data within its own disciplines but also helps identify overlaps and connections to the specialist domains of other consortia. Therefore, when integrating data from repositories and portals, a focus is placed on ensuring the interoperability to other knowledge graphs. This facilitates cross-disciplinary connections to and synergies with, for example, the NFDI4Memory Knowledge Graph, which is currently under development. The metadata for the research resources play a crucial role in the implementation of these interfaces. The Teams Knowledge Graph and Culture Information Portal (Dev) therefore are not only working to seamlessly integrate metadata into the knowledge graph and to develop and optimise harvesting and transformation processes, but are also analysing content overlaps.

Some workflows, amongst others, for the Culture Graph Interchange Format (CGIF) or the LIDO (Lightweight Information Describing Objects) format have already been implemented. A key milestone for the initial steps in testing further integration processes was the NFDI4Culture and NFDI4Memory workshop and hackathon Connecting Research Data from History and Musicology held at the German Historical Institute in Rome (DHI Rome) (29.-30.04.2024) earlier this year. Another focus of the workshop was to identify collections that can serve as interfaces to both consortia.

Participants of the two-day workshop were developers from the Team Knowledge Graph and the Team Culture Information Portal (Dev), data managers, librarians as well as scholars from the field of musicology and history. The introductory talks focused on the knowledge graph infrastructure applied both in NFDI4Culture and NFDI4Memory, as well as on the Culture Information Portal to ensure a basic understanding of the technical aspects and terms. This was followed by presentations by the DHI Rome team on the respective collections and research projects.

During the workshop, the following datasets from the DHI Rome were selected as test cases: the Libretti Collection from the Music History Department as well as letters from the digital edition Ferdinand Gregorovius: Poetry and Science. Collection of German and Italian Letters.

Based on these datasets further processes were tested, including those for standardised data in MARC21 (Libretti-Collection) as well as for the transformation and harvesting of data via a custom JSON-API (Gregorovius edition). Additionally, preparatory work was undertaken to develop a generalised integration routine for data via CMIF (Correspondence Metadata Interchange Format) and TEI XML. These transformation processes will soon be incorporated into the workflows. In addition, a Jupyter notebook was developed during the workshop that can be used as an example of a simple harvesting and transformation routine. The Jupyter notebook is available for reuse at the following link.

The NFDI4Culture and NFDI4Memory at the DHI Rome workshop served as a use case and demonstrator for interdisciplinary research data interoperability, and resulted in the inclusion of cross-domain research data into the Culture Knowledge Graph. The integrated metadata of the DHI Rome not only establish cross-consortium cross-links, but enrich the Culture Knowledge Graph by already showing the first connections to further musicological data sets within NFDI4Culture (see figure above).