Answers are grounded in this paper and cite their sources.
This poster was presented at RDA P27. Data peer review is fundamental to the trustworthiness of research datasets, yet remains one of the least standardised aspects of scholarly communication. Data journals are increasingly expected to engage with peer-review of datasets as well as of data papers. Data peer-review has been advocated as a highly desirable and beneficial practice (https://arxiv.org/pdf/1810.13040). Efforts in this direction have produced useful checklists (https://zenodo.org/records/11243822). Few FAIR assessment frameworks have been operationalised into reproducible workflows bridging dataset inspection and manuscript evaluation. However, current approaches to implement data peer-review in data journals’ editorial workflows are rare and not systematic, varying widely across journals and repositories, with limited consensus on criteria, methods, or reviewer support tools. At the Journal of Open Humanities Data (https://openhumanitiesdata.metajnl.com/), we are exploring how data peer-review might be integrated into our editorial workflow to enhance reusability and quality of the datasets, and contribute to community standards. In this poster, we consider several options for data peer-review integration, including documentation-focussed review, structured dataset sampling and automated validation against standards like FAIR and CARE. We report on early experiments developing a modular, human-led LLM-assisted framework via layered skill-based rubrics for dataset triage, provenance auditing, and reuse potential. Human reviewers retain editorial control while LLMs support systematic checking against documented criteria, cross-referencing data profiles, and drafting assessments for human refinement.
No takes yet. Share an insight, caveat, or question.
Farina et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: