Metadata mappings and transformations for PUNCH4NFDI Data Portal

The PUNCH4NFDI consortium brings together representatives from astroparticle, astro, particle, hadron and nuclear physics in order to improve and further develop research data infrastructures in alignment with the special requirements of researchers in these data-intensive research fields. The central element of this vision, the PUNCH Science Data Platform (PUNCH-SDP) is being developed to address needs such as discovery, reuse, and sharing of physicists' research results in digital form, leveraging the resources of federated infrastructures. The central element of the PUNCH-SDP, is the PUNCH4NFDI Data Portal. It provides access to a knowledge fabric connecting the elements of PUNCH4NFDI innovative Digital/Dynamic Research Products (DRP). This concept abstracts a wide range of digital research outputs, such as data in various states of processing from raw data to fully curated and published data collections, simulations, research papers, software, analysis workflows, etc., supporting data discovery, interoperability and reproducibility of research. Efficient support for this approach requires the development of a systematic and scalable approach to metadata curation in the system, focusing on operational use of heterogeneous metadata schemas defined in the PUNCH communities, the DRP metadata and broader data curation standards such as those defined in NFDI and EOSC. Our solution involves the development of a multi-level metadata schema that consistently extends the DataCite standard through the DRP level to the discipline-specific metadata standards used in PUNCH communities. Implementing this approach involves the creation of mappings and transformations between each of the schema levels, or provide the means to combine data within clearly defined execution workflows, as well as documenting, standardizing, and assuring quality of metadata records. By enabling structured and standardized metadata mappings across DRPs, our approach contributes to measurable improvements in data reusability and workflow reproducibility by making it easier to identify compatible datasets, reconstruct processing pipelines, and link research outputs across platforms. This contribution will highlight recent progress and outline next steps toward implementing metadata mappings and transformations within the PUNCH data portal.