International workshop at the University of Luxembourg | 06.03.2026

This workshop, organized by Peter Gilles, a professor of Luxembourgish linguistics and the creator of LuxASR, brings together international experts working on language technology and digital tools for small languages and varieties, with a focus on Frisian, Limbourgish, and Luxemburgish. The workshop aims to promote an exchange among experts regarding the unique challenges of developing language technology in small language contexts, as well as the potential of small language natural language processing (NLP) to enhance language technology development. Specifically, this involves improving the management of linguistic variation in models, preserving cultural diversity in the digital age, and offering innovative, tailored technical solutions for small language communities, such as AI text generation, automatic translation, and speech recognition.

Guiding questions for contributions

• What are the most important digital achievements for your language/speech community so far and who actually benefits from them?
• Which digital tools, resources or techniques are most urgently needed in your context?
• Which constraints are characteristic for small languages?
• How can AI be used productively and safely for small languages, especially when engaging with Big Tech?
• What kinds of partnerships (public institutions, NGOs, companies) are useful, and how can they respect GDPR, copyright and community ownership of language data?
• How can applied, community-oriented digital work be balanced with general academic research requirements?
• In which way are digital tools enhancing academic research through crowd-sourcing and citizen science?
• What would sustainable digital support for your language look like in 5–10 years and what needs to happen to achieve it?

Schedule

10:00–10:15 | Welcome & Introduction

10:15–10:45 | Nanna H. Hilton & Hans Van de Velde (Leeuwarden) | Digital Tools for Frisian: The Bird’s-Eye View

10:45–11:15 | Eduard Drenth (Leeuwarden) | Overview of ICT at the Fryske Akademie

11:15–11:45 | Anne-Marie Lutgen, Emilia Milano, Alistair Plum & Christoph Purschke (Luxembourg) | TRAVOLTA: Probing the Sociocultural Complexity of Luxembourgish through NLP

11:45–13:15 | Lunch Break & Campus walk

13:15–13:45 | Wilbert Heeringa (Leeuwarden, online) | Developing a UDPipe Model via Cross-Lingual Projection: The Case of Frisian

13:45–14:15 | Andreas Simons (Maastricht) | Limburgish Corpus: from decentralized dialects to digital tools

14:15–14:45 | Christopher Morse (Zenter fir d’Lëtzebuerger Sprooch) | D’Sproochmaschinn and Beyond: A Digital Corpus Infrastructure

14:45–15:15 | Léopold Hillah (Luxembourg) | LuxASR & real-time ASR

15:15–15:45 | Coffee Break

15:45–16:15 | Nina Hosseini-Kivanani (Luxembourg) | Finding the Luxembourgish Voice. Emotion, Charisma, and Technology

16:15–16:45 | Nils Rehlinger (Luxembourg) | LuxMT - Automatic Translation for Luxembourgish

16:45–17:15 | Peter Gilles (Luxembourg) | ASR ⇒ G2P ⇒ MFA ⇒ new-FAVE ⇒ R – A sociophonetic pipeline for messy crowd-sourcing data

17:15–18:00 | General Discussion: Needs, Strategies and Next Steps