Conference programme
The first version of the LaTeLL 2026 proceedings can be downloaded here.
International Conference on Language Technologies for Low-Resource Languages
28-30 September 2026 | Fes, Morocco
| 28-29 SEP Two technical days | 47 PAPERS 30 regular + 17 short | SINGLE TRACK 09:00-17:00 |
Presentation format
Regular papers 20 minutes: 15-minute presentation + 5-minute Q&A
Short papers 10 minutes: 7-minute presentation + 3-minute Q&A
Day 1 | Monday, 28 September 2026
09:15-09:30 Opening and welcome remarks: Prof R Mitkov, Dr Saad Ezzini and Dr Salima Lamsiyah
09:30-10:30 Keynote address – Prof. Nizar Habash | NYU Abu Dhabi, UAE
From Low-Resource Languages to Resource-Aware Language Technology
Keynote speaker introduction: Ruslan Mitkov and Salima Lamsiyah
10:30-10:50 Coffee break
Session 1 | Corpora, Annotation and Resource Development | 10:50-11:50
Session Chair: Yoan Gutiérrez
| Time | Format | Presentation |
| 10:50-11:10 | REGULAR 20 min | A Kazakh–Russian Corpus of Child-Directed Language for Low-Resource Languages Albina Mukusheva, Achille Fusco and Cristiano Chesi |
| 11:10-11:30 | REGULAR 20 min | Data Curation, Annotation Quality, and Error Patterns in Old English Automatic Lemmatisation Javier Martín Arista |
| 11:30-11:40 | SHORT 10 min | Creating the RozMuz Corpus: Applying Ethics and Technology in Language Research Alicja Helena Derych, Bartłomiej Alberski, Hubert Jankowski and Paweł Dembowski |
| 11:40-11:50 | SHORT 10 min | Towards a Universal Dependencies Treebank for Amazigh: A Tarifit Pilot Study Azzeddine Afrouni, Fadoua Ataa Allah and Jamal Abarnous |
Session 2 | Machine Translation and Cross-Lingual Transfer | 11:50-12:30
Session Chair: Salima Lamsiyah
| Time | Format | Presentation |
| 11:50-12:10 | REGULAR 20 min | Arabic Dialect-to-MSA Translation: A Comparative Evaluation of Large Language Models and Neural Machine Translation Maram I. Alharbi, Jihad R’BAITI and Ruslan Mitkov |
| 12:10-12:20 | SHORT 10 min | Cross-Lingual Transfer from Portuguese to Nheengatu: Evidence for Contact-Induced Convergence as a Computational Bridge Rafael Macario Fernandes |
| 12:20-12:30 | SHORT 10 min | Retrieval-Augmented Translation for Bohairic Coptic: A Pilot Benchmark and Hallucination Analysis So Miyagawa |
12:30-14:20 Lunch break
Session 3 | Morphology, Syntax and Formal Modelling | 14:20-15:30
Session Chair: Javier Martín Arista
| Time | Format | Presentation |
| 14:20-14:40 | REGULAR 20 min | Automating Mwotlap morphology using formal grammar Fabio Meroni, Alexandre François and Max Silberztein |
| 14:40-15:00 | REGULAR 20 min | Dependency Parsing Across the Resource Spectrum: Evaluating Architectures on High and Low-Resource Languages Kevin Guan, Happy Buzaaba and Christiane Fellbaum |
| 15:00-15:20 | REGULAR 20 min | A finite-state model of Innu verbal inflection Loïc Daignault-Pichette and François Lareau |
| 15:20-15:30 | SHORT 10 min | Morphological Operations in the Persian Verbal System Marzieh RABIEI and Max Silberztein |
15:30-15:50 Coffee break
Session 4 | Training Data and Model Adaptation | 15:50-17:10
Session Chair: Saad Ezzini
| Time | Format | Presentation |
| 15:50-16:10 | REGULAR 20 min | A Massive Open-Source Corpus for Valencian: Over 4.7 Billion Tokens for Low-Resource Language Modelling Yoan Gutiérrez, Juan Pablo Consuegra‐Ayala, Robiert Sepúlveda-Torres and Rafael Muñoz Guillena |
| 16:10-16:30 | REGULAR 20 min | LuxIT: A Luxembourgish Instruction Tuning Dataset from Monolingual Seed Data Julian Valline, Cedric Lothritz, Siwen Guo and Jordi Cabot Sagrera |
| 16:30-16:50 | REGULAR 20 min | Topic Modeling for Moroccan Darija: A Comparative Study of Classical Machine Learning Approaches Salma Mekaoui, Ilham Chaker, Arsalane Zarghili and Nikola S. Nikolov |
| 16:50-17:10 | REGULAR 20 min | Distil and Evolve: Compact Adaptive Document Classification with Continual Learning for Low-Resource Settings Mehdi Soufiane, Mourad Hammale, Kawtar Zerhouni, Adil Ahidar-Coutrix and Rachid Benouini |
Day 2 | Tuesday, 29 September 2026
Session 5 | Language Model Behaviour and Evaluation | 09:30-10:30
Session Chair: Ernesto Luis Estevanell
| Time | Format | Presentation |
| 09:00-09:20 | REGULAR 20 min | Do Large Language Models Style-Shift? Register-Conditioned Stylistic Fronting in AI-Generated Icelandic Anton Karl Ingason, Johanna Mechler and Lilja Björk Stefánsdóttir |
| 09:20-09:40 | REGULAR 20 min | Benchmarking Large Language Models on Mandarin Proverb Explanation and Contextual Matching Xiaojing Zhao, Salima Lamsiyah, Emmanuele Chersoni and Han Xu |
| 09:40-10:00 | REGULAR 20 min | LUXDIAG-RAG: Diagnostic Evaluation of Retrieval-Augmented Generation for Luxembourgish Reading Comprehension Keerthana Murugaraj, Hedi Tebourbi, Christophe Friezas Goncalves and Salima Lamsiyah |
| 10:00-10:30 | Sponsor presentation | Diamond Sponsor presentation — AutoTrust AI Endless Memory is the Key of Recursive of Self-Improvement Daniel (Xunzhu Tang), AutoTrust AI Sponsor Introduction: Saad Ezzini |
10:30-10:50 Coffee break
Session 6 | Human-Centred NLP and Education | 10:50-11:50
Session Chair: Salima Lamsiyah
| Time | Format | Presentation |
| 10:50-11:10 | REGULAR 20 min | Assessment of Human-in-the-Loop Multi-LLMs for Low-Resource Educational Data Expansion Christophe Friezas Gonçalves, Hedi Tebourbi, Christoph Schommer and Salima Lamsiyah |
| 11:10-11:30 | REGULAR 20 min | Human-Centred Approaches in Low-Resource Languages for Educational Applications and Language Learning: A Systematic Literature Review Noura EL MOUSSA |
| 11:30-11:50 | REGULAR 20 min | SimuLe-Lux: A Neuro-Symbolic Teaching System for Diagnosing L2 Reading Comprehension Christophe Friezas Gonçalves, Christoph Schommer and Salima Lamsiyah |
Session 7 | Arabic and Dialect NLP | 11:50-12:50
Session Chair: Salmane Chafik
| Time | Format | Presentation |
| 11:50-12:10 | REGULAR 20 min | Towards Readability Assessment for Under-Resourced Arabic Dialects: A Study of Moroccan Darija Houdaifa Atou, Nouran Khallaf, Salima Lamsiyah and Ruslan Mitkov |
| 12:10-12:30 | REGULAR 20 min | Why Current XAI Is Not Enough for Arabic NLP? A Critical Survey of the Explainability Gap Salima Lamsiyah and Ruslan Mitkov |
| 12:30-12:40 | SHORT 10 min | Benchmarking Speech Foundation Models and LLMs for Sentiment Analysis in Low-Resource Najdi Arabic of Saudi Arabia Nadia Ghezaiel, Maram I. Alharbi and Ruslan Mitkov |
| 12:40-12:50 | SHORT 10 min | Spatial Entity Extraction Methods from Arabic Texts in the Context of Epidemiological Surveillance: A Comparative Study Fatima Ezzahra El Houbri, Najlae IDRISSI, Mathieu Roche and Sarah Valentin |
12:50-14:20 Lunch break
Session 8 | Speech and Spoken-Language Technologies | 14:20-15:20
Session Chair: Nadia Ghezaiel
| Time | Format | Presentation |
| 14:20-14:40 | REGULAR 20 min | Building an Aligned Speech Corpus for Northern Russian Kseniia Protonina and Daniil Ignatev |
| 14:40-15:00 | REGULAR 20 min | Probing Geographic Performance Gaps in Moroccan ASR with ALAMA: Annotating Local Audio for Moroccan Arabic Avery Cole Kanel, Christian Schuler, Bouazza Laracha, Yassine Chaouri, Imrane Lbouhli, Yusser Al Ghussin and Timo Baumann |
| 15:00-15:20 | REGULAR 20 min | Generator-Guided Amount Recovery for Voice-Based Financial Record-Keeping in Mooré-French Code-Switched Speech Maimouna Ouattara, El-Hacen Diallo, Fred Philippy, Abdoul Kader Kaboré, Jacques Klein and Tegawendé F. Bissyandé |
15:20-15:50 Coffee break
Session 9 | Session 9 | Retrieval, RAG and Document Processing | 15:50-16:40
Session Chair: Houdiaifa Atou
| Time | Format | Presentation |
| 15:50-16:10 | REGULAR 20 min | LëtzCross: A Cross-Lingual Page-Level Benchmark for Multimodal Retrieval over Luxembourgish Documents Omar El Bachyr, Fred Philippy, Laura Maria Bernardy, Saad Ezzini, Jacques Klein and Tegawendé F. Bissyandé |
| 16:10-16:30 | REGULAR 20 min | Improving OCR for a Latvian Pronunciation Dictionary Viesturs Jūlijs Lasmanis |
| 16:30-16:40 | SHORT 10 min | Agentic retrieval for low-resource languages: An exploratory system design Malithi P. Alahapperuma, Andreas Vlachidis and Antonis Bikakis |
| 16:40-16:50 | SHORT 10 min | Clinical Entity Recognition from Electronic Health Records and Linking to Biomedical Knowledge Bases in Low-resource Language Settings Eno-Martin Lotman, David Hübner, Kris Collins, Svetla Boytcheva and Ivelina Nikolova-Koleva |
| 16:50-17:00 | SHORT 10 min | Beyond RAG: A Multi-Graph, Multi-Agent, Recursive Retrieval Architecture for Traceable and Source-Grounded Legal Answers Assia Bouamir, Marie Bonnin, Youssef Al Mouatamid and Jihad Zahir |
17:00-18:00 Closing session, awards and announcements
Website | Video Presentations
| Format | Video Presentation |
| REGULAR 20 min | FLICK: Few-Label Incremental Learning for Low-Resource Dialects Ali Almutairi, Abdullah Alsuhaibani, Shoaib Jameel, Aditya Joshi, Gelareh Mohammadi and Imran Razzak |
| REGULAR 20 min | Linguistic Proximity Enables Pivot-Based Machine Translation for an Under-Resourced Tribal Language Pooja Singh, M Kaab Bin Shahid, Atai Waris Khan, Aryan Kumar Jha and Sandeep Kumar |
| REGULAR 20 min | Cross-Lingual Hate Speech Detection in Low-Resource Languages: The Case of Persian and Kurdish Shahin Yousefi, Ernesto Luis Estevanell-Valladares and Ruslan Mitkov |
| REGULAR 20 min | Is Overconfidence Language-Specific? Cross-Lingual Calibration and Recalibration Transfer in Base and Instruction-Tuned LLMs Divya Gupta and Ruslan Mitkov |
| REGULAR 20 min | Knowledge Tracing for Early Childhood Learners with Minimal Telemetry from Low-Connectivity Environments: Evidence from India’s Anganwadi Ecosystem Badmavasan Kirouchenassamy, Rahul Singh and Sudeep Gowrishankar |
| REGULAR 20 min | Kuwain 1.5B: An Arabic SLM via Language Injection Khalil Hennara, Sara Chrouf, Mohamed Motasim Hamed, Zeina Aldallal, Mohammad Omar Hadid and Safwan AlModhayan |
| REGULAR 20 min | A Neuro-Symbolic RAG System for Marine Environmental Law: from Domain Ontology to Knowledge Graphs and Context-Enriched Legal Question Answering Hadiza Dite Maa Assoumana Souley, Marie Bonnin, Youssef Al Mouatamid and Jihad Zahir |
| SHORT 10 min | Evaluating Model-Task Fit in Arabic Word Sense Disambiguation Yousef Younes, Abdelhalim Hafedh Dahou and Brigitte Mathiak |
| SHORT 10 min | Enhancing Urdu ASR with Whisper v3: Fine-Tuning on Latest Datasets and Realistic Multi-Speaker Evaluation with SLM Post-Processing Zehra Ahmed, Farah Inayat, Zuha Aqib and Sajjad Haider |
| SHORT 10 min | Baseer: A Vision-Language Model for Arabic Document-to-Markdown OCR Khalil Hennara, Muhammad Hreden, Mohamed Motasim Hamed, Ahmad Bastati, Zeina Aldallal, Sara Chrouf and Safwan AlModhayan |
| SHORT 10 min | Worldview Annotation for Low-Resource Languages Tetiana Ilman |
| SHORT 10 min | A Multilingual Sentence-Transformer Baseline for Serbian Newspaper Topic Classification Sasa Petalinkar, Milica Ikonić Nešić, Ranka Stankovic and Jelena Graovac |
| SHORT 10 min | Wasm: A Pipeline for Constructing Structured Arabic Interleaved Multimodal Corpora Khalil Hennara, Ahmad Bastati, Muhammad Hreden, Mohamed Motasim Hamed, Zeina Aldallal, Sara Chrouf and Safwan AlModhayan |
| SHORT 10 min | Linguistic Specialization of Arabic in Text Embedding Models via Multi-Teacher Knowledge Distillation Ahmad Abdelfattah, Zeina Aldallal, Sara Chrouf and Safwan AlModhayan |