Parallel Session 12 – Responsible Research Data Management II
Date: Friday, 3 July 2026, from 09:00 to 10:30
Moderator: TBC
Location: R5
12.1) AI 4 RDM: Exploring AI for Research Data Management Support at KU Leuven Libraries
Presenter: Miara Fraikin, KU Leuven Libraries, Belgium
At KU Leuven, Research Data Management Support is organised as a collaborative network, coordinated by the RDM Competence Centre. Within this network, the library plays an important role: its staff provide training for researchers, curate comprehensive RDM webpages, assist via an RDM helpdesk, manage the institutional data repository, organise RDM-related events, review data management plans, and the library heads the RDM Competence Centre. In addition, we have been exploring how AI can enhance these RDM support services. This presentation will share our approach, progress, and key insights from the AI4RDM project.
While researchers are increasingly aware of the importance of sound research data management, RDM is still often viewed as an administrative burden taking away valuable research time. It also appears that researchers often struggle to find the right support channels, resulting in only a small proportion of our 14,000 researchers and 7,000 PhDs accessing the library’s RDM support staff and materials. Recognizing AI’s potential to deliver personalized information and improve scalability, we have begun exploring how AI can be used to enhance our RDM support services since September 2024.
In this presentation, we will detail four (pilot) projects, each focusing on a specific aspect of AI integration within RDM-support workflows. Our first project, in collaboration with the Research Coordination Office and the context of a master thesis in the advanced master programme of Artificial Intelligence, explored using AI to automate the review of Data Management Plans (DMPs). Rather than manually reviewing all 700 DMPs each year, we aimed to improve efficiency by automatically categorizing them and only manually reviewing those needing feedback. A key finding was the need for structured domain knowledge. This knowledge, currently distributed across various unstructured sources, was at risk of no longer being compliant with the recently updated KU Leuven RDM policy. Since manually reviewing all materials would be resource-intensive, our second project, in partnership with the master of digital humanities, uses AI to identifyconflicting documentation.
While the first two projects focus on improving internal RDM support workflows and documentation, the third and fourth projects are intended to benefit researchers more directly. The third project involves developing an AI agent with access to our RDM documentation and evaluating its potential as a chatbot on the KU Leuven RDM website to further strengthen researchers’ RDM self-reliance. The fourth project examines researchers’ workflows, assessing which AI tools are currently used for research data management, evaluating their compliance with KU Leuven regulations, and identifying further needs that could be addressed with tailored solutions.
In summary, this presentation explores the potential use of artificial intelligence within Research Data Management support at KU Leuven. By sharing our experiences across four distinct projects, we aim to highlight the practical opportunities and challenges of integrating AI into RDM workflows. We hope our findings will contribute to ongoing discussions about AI’s role in academic support services and offer useful insights for institutions facing similar challenges.
12.2) Human-Centred AI for Data Management Plan Assessment and Support
Presenter: Anne-Marie Tuikka, University of Turku, Finland
Research libraries are increasingly expected to safeguard responsible research data practices under conditions of technological uncertainty, rising administrative burden, and constrained resources. In this context, a central question is whether generative artificial intelligence can reliably support expert-led research data governance without undermining quality, accountability, or trust. This presentation addresses that question by empirically examining how AI-generated assessment of Data Management Plans (DMPs) can support rather than replace professional judgement in research data management.
We present a study of DMP Analyser, a prototype AI tool developed by the authors to support expert review of DMPs using established evaluation criteria. The system applies a modular analysis pipeline to produce structured and traceable outputs and explicit audit trail from each judgement to sentence-level evidence in the original DMP. This design enables experts to inspect, validate, and contest AI outputs, positioning the tool as a decision-support system rather than an automated evaluator. The analysis pipeline is configurable and can utilise different large language models and evaluation rubrics, enabling institutions to apply shared criteria consistently and to accumulate an institutional knowledge base grounded in expert-validated DMP assessments.
The study analyses a corpus of DMPs submitted to the 2025 Research Council of Finland funding call from Tampere University and the University of Turku. All funded projects are required to submit a DMP, which is reviewed by institutional data support professionals who provide written feedback based on a shared evaluation rubric. This setting offers a realistic governance context in which expert judgement is formally embedded and documented.
Following expert assessment, the same DMPs are independently analysed using the DMP Analyser between January and May 2026, which produces structured evaluations aligned with the criteria used by human reviewers. AI-generated assessments were then systematically reviewed by the research team and compared against expert feedback and the original DMP texts.
To support this analysis, the DMP Analyser was extended to store expert comments, AI outputs, and validation decisions, enabling the calculation of quantitative indicators of agreement and divergence across individual assessment criteria. This allows us to examine how the reliability and usefulness of AI support vary by topic and type of judgement, and to identify where human oversight remains essential.
The findings highlight both the potential and the limitations of AI-assisted DMP support. While AI performs well on structurally explicit criteria, it struggles with context-sensitive judgements and ambiguous descriptions. Rather than treating this as a benchmark of AI performance, the study demonstrates a human-centredapproach in which such limitations are made visible, traceable, and actionable for expert reviewers. The contribution lies in showing how generative AI can function as a transparent and verifiable decision-support infrastructure that strengthens professional judgement and institutional responsibility, rather than replacing them.
12.3) Librarians in the Loop: Hybrid AI for Reliable Data Extraction from Historical Loan Registers
Presenter: Christopher Kermorvant, TEKLIA, France
The PRET19 project, led by the Bibliothèque interuniversitaire de la Sorbonne (BIS) and the Université Paris-Est Créteil in partnership with other libraries and TEKLIA, converts 19th-century handwritten loan registers into a reusable dataset on the history of reading, academic disciplines, and higher education. Rather than relying on a single “black box” AI system, PRET19 employs a controlled, stepwise approach in which various tools are integrated and librarians verify each stage, from identifying borrower entries to connecting books with the national Sudoc catalogue. When applied to five registers from 1850 to 1893, this process reconstructs tens of thousands of individual loans and several thousand borrowers, with the majority of loans successfully matched to at least one catalogue record. PRET19 demonstrates how research libraries can responsibly utilise AI to re-engage with loan registers as a vital component of their documentary heritage, as well as for analysing the circulation of books and ideas.