The HMC Dashboard on Open and FAIR Data - Support and Insights for the Helmholtz Data Landscape

Dashboards have become an important tool to monitor and assess the state of Open Science for research performing organisations [see [1]–[4]]. In 2022 the Helmholtz Metadata Collaboration launched the HMC Dashboard on Open and FAIR Data which has become an integral part of the Helmholtz data landscape with 17 of the 18 Helmholtz centers connected to the dashboard. Here, we report on details of a major upgrade in December 2024 and the subsequent impact of these changes on the Helmholtz community. We focus in particular on the dashboard consulting services prototyped in our previous publication [5]. The dashboard harvests literature metadata from the libraries of the Helmholtz research centers. Data publications linked to the harvested literature are then identified using the SCHOLIX API and assessed using F-UJI [6]. The information collected is presented in an interactive dashboard. It allows users to explore in which repositories Helmholtz researchers make their data publicly available, to engage Helmholtz communities, and to identify gaps towards improving the FAIRness of Helmholtz data. The dashboard is publicly available on https://fairdashboard.helmholtz-metadaten.de. In December 2024 we published a major update to the dashboard, launching version 2.0.0 (see section Resources). Updates included a major refactoring of the dashboard harvesting infrastructure so that it now runs weekly. Because of the regular harvesting, problems in Helmholtz data infrastructure are flagged and resolved quickly ensuring the findability and visibility of Helmholtz outputs is maintained. Since launching the update we have identified and solved problems with infrastructure in 9 of the 17 Helmholtz centers connected to the dashboard. One example relates to the Marc-21 metadata schema. In less than a week we were able to identify that this resulted in export errors for infrastructure in multiple Helmholtz centers affecting their visibility. We contacted the affected infrastructures immediately and supported them in resolving the problem by acting as a central liaison point for MARC-21. The result is that there was minimal disruption to the visibility of Helmholtz records in higher level (non-Helmholtz) harvesting infrastructure. Such counselling is indicative and illustrative of the power of the dashboard to provide timely feedback to Helmholtz infrastructure. Reusability of the dashboard program code was enhanced in the update. The source code repository is now OpenSSF [7] and REUSE [8] compliant. We welcome users of either the code or approach by all research communities. How we collect data: We built the HMC Toolbox in Python3 [9] that collects metadata of literature and data publications by the Helmholtz community. For literature metadata this is done via the OAI Protocol for Metadata Harvesting (OAI-PMH) [10] using the well known library Sickle: OAI-PMH for Humans [11]. The update now includes validation to check whether publications labeled as literature are correctly classified. The connection between literature publications and their related data publications is done via the external tool ScholeXplorer [12], which finds related data publications for literature publications by PIDs and offers structured metadata about them.