Metadata for DiscoRSE: An Open Educational Resource Catalogue for Research Software Engineers

Abstract

Research Software Engineers (RSEs) work at the intersection of research practice, software engineering, and domain-specific knowledge — a combination that few people enter the field with. The RSE community has responded to this skills gap by producing a large body of Open Educational Resources (OERs), but these resources are often scattered across institutional websites, poorly described, and hard to reuse beyond the institutions that created them. DiscoRSE addresses this gap by aggregating, curating, and enriching OERs for RSEs from multiple sources, with a shared metadata schema at its core to make them more findable, accessible, interoperable, and reusable (FAIR). In this paper, we report on the metadata work for DiscoRSE: using schema:LearningResource as the reference, we performed a property-level crosswalk against five existing schemas for OERs and training materials — the Bioschemas TrainingMaterial profile, the Allgemeines Metadatenprofil für Bildungsressourcen (AMB), the DINI/nestor Metadata Schema for RDM Training Materials, Skills4EOSC, and the MoDALIA ontology used in DALIA. The analysis identifies a small core of properties consistently supported across all five targets, a wider set supported by four targets, five candidate additions not yet covered by LearningResource, and a set of open discussion points for the community. On this basis we propose an initial core profile for DiscoRSE and outline next steps, i. e., ETL pipelines, AI-assisted metadata generation, and selection of an OER platform.