On Tuesday, 9 December we will be hosting three pre-conference workshops. Participation is including with the registration fee. Space is limited, so please select the workshop you are interested in during the registration process. Information about the three workshops are available below.

Impresso Datalab: Exploring and annotating multilingual digitised newspapers and radio archives

Organizers: Caio Mello and Marten Düring

This three-hour hackathon-style workshop offers hands-on experience with the Impresso Datalab, a platform that provides access to multilingual historical newspaper and radio archives along with machine learning models for semantic enrichment tasks like named entity recognition, OCR quality assessment, and press agency detection. Participants will work in small groups to experiment with linking their own research datasets (text and images) to the Impresso corpus across different languages, time periods, and modalities using Jupyter notebooks. The workshop is designed for all members of the digital humanities community—coding experience is helpful but not required—and participants should bring a laptop and, ideally, a research dataset they’d like to work with (which organizers may preprocess beforehand). The session will include a platform demo, group work defining research objectives and experimenting with notebooks, and concluding with group presentations. This workshop would appeal to researchers interested in computational approaches to cultural heritage data, especially those working with historical media collections or seeking to overcome the limitations of standard graphical interfaces for digital archives.

Text-Based Emotion Detection (TBED) in Historical Letters

Organizers: Albina Sarymsakova and Patricia Martín-Rodilla

This workshop introduces text-based emotion detection (TBED) methods applied to historical sources, specifically 16th-century correspondence written by royalties and aristocracies. Through four sessions combining theory and hands-on practice, participants will learn emotion classification frameworks (discrete vs. dimensional models), explore NLP techniques and Python-based tools for emotion detection, annotate sample letters from the Corpus of Early English Correspondence Sampler (CEECS), and evaluate large language models on historically annotated texts. The workshop takes an interdisciplinary approach integrating linguistic, historical, and computational perspectives, making it valuable for researchers in linguistics, history, digital humanities, computer science, and computational linguistics. Participants should have basic Python proficiency and bring a laptop with Python 3.7+, Jupyter Notebook, and required NLP libraries installed (setup guide provided in advance). By the end, attendees will understand emotion classification frameworks, be familiar with core NLP techniques for TBED, know how to interpret historical correspondence within emotional contexts, and gain practical skills in applying computational tools to analyze emotions in early modern texts.

Computational Notebooks for Digital Humanities

Organizers: Mattia Bunel, Elisabeth Guerard, Raphaelle Krummeich, Hugues Pecout and Sebastien Rey-Coyrehourcq

This full-day workshop explores the use of computational notebooks (Jupyter/Python) in digital humanities research, focusing on literate programming principles, reproducibility, and academic publishing standards. Organized by an interdisciplinary Notebook Working Group from French research institutions, the workshop addresses how notebooks can contribute to the scientific process through both theoretical discussions and hands-on practice. The morning session provides epistemological and technical introduction to notebooks, followed by a guided tutorial where participants create and publish replicable notebooks with trainer support. The afternoon addresses scientific reproducibility challenges, examines how Git and associated platforms (GitHub, GitLab) have become academic publishing standards, and features presentations from two computational journals—the Journal of Digital History and RZine—on their editorial workflows and reproducibility practices. Limited to 30 participants, the workshop is designed to be accessible without requiring participants to write code themselves, though attendees should be comfortable reading source code and have basic familiarity with semantic web technologies like SPARQL queries (introductory resources provided for those unfamiliar). Participants need only a laptop with good internet connection to access the Jupyter Lab instance set up for the event via web browser, making notebook methodologies accessible to humanities researchers while bridging theory and practice.