Jérôme Darmont – 30th European Conference on Advances in Databases and Information Systems (ADBIS) 2026, Orléans

ADBIS and I.

Dear all,

Unfortunately, I can’t be with you this evening. It’s a big frustration for me, but you may have met some of my PhD students this week!

My first ADBIS conference was in 2006 in Thessaloniki. I remember a lovely hotel with a nice garden. But above all, I met a great, warm community there, with Yannis as a conductor, uh, General chair (now ADBIS Chair).

Somehow, it took me six years to get back to ADBIS, in Poznan. Next, I let time flew again until Poitiers (France). Unfortunately, I couldn’t attend. Another frustration. Then, at a moment in my career, I wanted some changes. It happened at ADBIS 2016 where, many new ideas have emerged and inspired me. Afterwards, I had a nice series in a rows: Prague, Nicosia, Budapest, Bled and Lyon.

ADBIS 2020 was particular for me, since I had the honor to be General chair this year. But you remember the Covid pandemic has occurred at this time, and we had to reorganize the conference online pretty quickly! It was another frustration, though the conference was a success.

Eventually, I attended ADBIS in Torino and Bayonne. By the way, for the younger of the audience, can you identify the countries from the cities I mentioned in my speech without using ChatGPT? ☺

Last but not least, I am extremely honored to share this « Test of time award » with my co-authors, Kamel Aouiche and Pierre-Emmanuel Jouve, whom were PhD student and post-doc at the time, respectively. The paper, “Clustering-based materialized view selection in data warehouses” was a nice collaboration between two research teams, and I’m delighted it is still cited after 20 years.

Now I’m done with these last words:
“Frustration? What frustration? It was only pleasure. Thank you so much, ADBIS!”

Partenariat Hubert Curien (PHC) Braccusi, France-Roumanie.

Ce projet vise à mettre en œuvre un système de détection de contenus préjudiciables, en utilisant une approche fondée sur les données. Les principaux objectifs sont les suivants : 1) développer des modèles efficaces de détection des contenus préjudiciables par apprentissage automatique et
entraîner selon une approche fondée sur les données ; 2) déterminer si les contenus préjudiciables transcendent les facteurs sociaux, culturels, politiques et économiques.

R. El-Idrissi, J. Agoun, J. Darmont, S. Loudcher, « Content Learning for Metadata Extraction and Enrichment of Historical Hand-Drawn Illustrations », New Trends in Databases and Information Systems – ADBIS 2026 Short Papers, Sept-October 2026; Communications in Computer and Information Science, Springer, Heidelberg, Germany.

Abstract: The digitisation of cultural heritage collections requires automated metadata generation to improve resource discovery and reuse. This paper presents the Content Learning for Metadata Extraction and Enrichment (CLEAD), a multimodal approach for generating metadata from illustrations in digitised archaeological diaries. CLEAD combines illustration detection and segmentation with vision-language and large language models to generate and enrich illustration descriptions using page-level textual context. Experiments on the DataLAC and IlluHisDoc datasets show that integrating visual and textual information improves the semantic quality and relevance of the generated metadata. These results demonstrate that CLEAD effectively supports the indexing and retrieval of visual heritage resources in digital archives and data lakes.

A. Diouan, S. Loudcher, J. Darmont, E. Ferey, « Discovering Relationships in Data Lakes Using Large Language Models: An Industrial Case ». 28th International Conference on Big Data Analytics and Knowledge Discovery (DaWaK 2026), Graz, Austria. LNCS.

Abstract: Data lakes rely on metadata to remain usable, yet this meta data is often limited or weakly informative for column relationship discovery, especially in ERP-derived datasets with coded or abbreviated schema labels. We propose ColRel, a two-stage method that builds column embeddings from metadata and data available at ingestion time. In difficult cases, such as coded schemata, business dictionaries help better interpret column names and support the generation of short natural-language descriptions used in the second stage. Experiments on public benchmarks and an industrial ERP dataset show that ColRel is particularly effective in semantically related, weak-signal settings.