By Jon Johnson & Alina Danciu
The 17th European DDI (EDDI 2025) User Conference was hosted from 1-5 December 2025 in Budapest, Hungary by the Research Documentation Centre at the ELTE Centre for Social Sciences .
As part of EDDI’s mission to build and strengthen the DDI community, the conference moves to a new city each year. By that measure, EDDI 2025 was a resounding success. One-third of all presentations were delivered by first-time speakers, and 40% of participants were attending EDDI for the first time. Twenty percent of the delegates came from the region, double the average of the previous four years. In total, the conference featured more than 50 presentations or sessions and nearly 87 delegates.
This strong regional participation was supported by the generous sponsorship of CESSDA, CDSP, SciencesPo, Making Sense, and the Metacurate-ML project, which enabled a larger number of local and early-career delegates to attend.
CESSDA’s sponsorship also made it possible to run a special workshop for small and under-resourced archives, a common challenge especially in Central and Eastern Europe. Participants highlighted the value of learning not only about technical infrastructure, but also about governance, staffing, and skills development. Many good connections were formed, and several collaborative initiatives are already being planned.
This year also marked a farewell to IZA, which closed at the end of 2025. IZA has been a long-standing supporter of EDDI, hosting the very first conference in Bonn in 2009 and providing the conference’s web presence for many years.
The conference opened with a welcome from Zsolt Boda, Director General of the ELTE Centre for Social Sciences, followed by the first keynote from Merce Crosas of the Barcelona Supercomputing Centre. Her talk set the tone for one of the conference’s central themes: the possibilities, limitations, and challenges of making research data “AI ready.”
Social science archives represent a uniquely rich playground for AI, holding diverse yet structured, machine-consumable historical data that can now be explored in ways unimaginable just a few years ago. At the same time, semantic diversity across languages, cultures, and disciplines remains a major challenge. Many presentations explored how Al and machine learning (ML) technologies can support classification, enhancement, and annotation as foundational steps toward AI-ready data. DDI standards — both structural (DDI-C, DDI-L, DDI-CDI) and semantic (CV’s, XKOS, SDTL) — together with allied standards such as ELSST, provide a firm basis for this work. Next year’s presentations will no doubt show how these limitations are being addressed in practice.
A new feature this year was a DDI Scientific Board’s “quick-fire” session, which provided a concise overview of DDI Alliance activities over the past year, an excellent introduction for new attendees.
Tuesday’s programme included sessions on FAIR data, with perspectives from the Czech Republic and EU-funded projects, training on questionnaires and open science, and two sessions from the METACURATE-ML project. The day concluded with an excellent conference dinner at the Astoria Hotel.
The second day opened with an insightful keynote from Béla Janky of ELTE on the imbalance in reward for data managers and why this is a problem for sustaining a high quality research culture.
Further sessions examined the use and adoption of DDI specifications within an increasingly heterogeneous standards landscape, alongside case studies from French health data infrastructures that are adopting an interoperability-first approach, raising new questions and opportunities for DDI.
The poster sessions again proved an important space for newer members of the community to engage in deeper discussions. Many posters focused on training, outreach, and simplifying metadata workflows. These were followed by a hands-on session to create DDI use cases for the DDI Alliance website, highlighting how metadata directly improves data quality and reuse.
The conference concluded with sessions on data-sharing infrastructures and a session showcasing a range of new software tools.
Presentations will be made available on Zenodo at https://zenodo.org/communities/eddi2025.
Looking ahead, EDDI 2026 will be held in Brussels and hosted by the Social Sciences and Digital Humanities Archive of Belgium (SODHA) (dates to be confirmed).
SODHA will also host an Archiving Research and Researching Archives Summer School from 4-7 May 2026 in Brussels, supported by EDDI. You can register your interest here: