|
|
|
| |||||||||||||||
DeTermIt 2026 : Workshop @ LREC 2026 - Evaluating Text Difficulty in a Multilingual Context | |||||||||||||||
| Link: https://determit2026.dei.unipd.it/ | |||||||||||||||
| |||||||||||||||
Call For Papers | |||||||||||||||
|
DeTermIt!
Workshop @ LREC 2026 Second Workshop on Evaluating Text Difficulty in a Multilingual Context Location: Palau de Congressos de Palma, Palma de Mallorca (Spain) ##################### First Call for Papers Schedule - Paper submissions: 23 February 2026 - Notification of acceptance: 13 March 2026 - Camera-ready due: 30 March 2026 - Workshop: one of 11, 12, or 16 May 2026 (half-day) All deadlines are 11:59PM UTC-12:00 AoE (“Anywhere on Earth”) For more information, please visit: Website: https://determit2026.dei.unipd.it/ ##################### In today’s interconnected world, where information dissemination knows no linguistic bounds, it is crucial to ensure that knowledge is accessible to diverse audiences, regardless of language proficiency and domain expertise. Automatic Text Simplification (ATS) and text difficulty assessment are central to this goal, especially in the age of Large Language Models (LLMs) and Generative AI (GenAI), which increasingly mediate access to information. The second edition of the DeTermIt! workshop focuses on the evaluation and modeling of text difficulty in multilingual, terminology-rich contexts, with a particular emphasis on the interaction between: - text simplification, - terminology and conceptual complexity, and - LLM/GenAI-based generation and rewriting. The 2026 edition builds on the first DeTermIt! workshop held at LREC-COLING 2024 (https://determit2024.dei.unipd.it/), as well as related initiatives such as the CLEF SimpleText track (https://simpletext-project.com/), which provides reusable data and benchmarks for scientific text summarization and simplification. DeTermIt! 2026 aims to bring together researchers and practitioners interested in terminology-aware simplification, lexical and conceptual difficulty, and evaluation protocols for GenAI systems. We welcome contributions that address theoretical, methodological, and applied aspects of text difficulty, including resource creation and evaluation (e.g., corpora, datasets, and benchmarks), with a focus on how linguistic complexity, specialized terminology, and domain knowledge interact with human understanding. In particular, we encourage work that explores how LLMs and GenAI can be evaluated, constrained, or guided to produce readable, faithful, and accessible texts. ##################### Topics of Interest ##################### We invite submissions on (but not limited to) the following themes: 1. Theoretical and Modeling Perspectives - Cognitive and linguistic models of text and lexical complexity. - Multilingual readability and text difficulty prediction. - Modeling conceptual difficulty and domain-specific terminology. - Theoretical connections between lexicography, terminology, and text simplification. 2. Terminology and Conceptual Complexity - Identification and classification of specialized terms and concepts. - Estimation of term difficulty for lay readers and second language learners. - Use of terminological databases, ontologies, and knowledge graphs in simplification pipelines. - Methods for adapting domain-specific terminology for accessible communication (e.g., in medicine, law, technology). 3. Generative and Explainable AI for Text Simplification - LLM- and GenAI-based approaches to text simplification and paraphrasing. - Terminology-Augmented Generation (TAG) and term-preserving simplification. - Evaluation of GenAI outputs: readability, factuality, terminology fidelity, and hallucination analysis. - Readability-controlled or difficulty-controlled generation; controllable simplification. - Human-centered and explainable approaches to text accessibility in GenAI systems. 4. Resources, Benchmarks, and Evaluation Frameworks - Corpora, annotation schemes, and benchmarks for text difficulty and simplification. - Datasets and methods for evaluating terminology-aware simplification and explanation. - FAIR and reusable resources for multilingual text accessibility. - Evaluation protocols and metrics for cross-lingual and cross-domain simplification and GenAI-based rewriting. 5. Applications and Case Studies - Domain-specific simplification (e.g., healthcare, legal, scientific communication). - Tools and systems for educational settings, language learning, or accessible communication. - User studies, human evaluation setups, and mixed-method approaches to assessing text difficulty and GenAI-assisted simplification. - Industrial and real-world experiences with integrating ATS and terminology into LLM-driven workflows. ##################### Submission Guidelines ##################### We invite original contributions, including research papers, case studies, negative results, and system demonstrations. When submitting a paper through the START system of LREC 2026, authors will be asked to provide essential information about language resources (in a broad sense: data, tools, services, standards, evaluation packages, etc.) that have been used for the work described in the paper or are a new result of the research. ELRA strongly encourages all authors to share the resources described in their papers to support reproducibility and reusability. Papers must be compliant with the stylesheet adopted for the LREC 2026 Proceedings (see https://lrec2026.info/authors-kit/). The workshop proceedings will be published in the LREC 2026 workshop proceedings. PAPER TYPES We accept three types of submissions: - Regular long papers – up to eight (8) pages of content, presenting substantial, original, completed, and unpublished work. - Short papers – up to four (4) pages of content, describing smaller focused contributions, work in progress, negative results, or system demonstrations. - Position papers – up to eight (8) pages of content, discussing key open challenges, methodological issues, and cross-disciplinary perspectives on text difficulty, terminology, and GenAI. References do not count toward the page limits. ##################### Organizers ##################### Chairs Giorgio Maria Di Nunzio, University of Padua, Italy Federica Vezzani, University of Padua, Italy Liana Ermakova, Université de Bretagne Occidentale, France Hosein Azarbonyad, Elsevier, The Netherlands Jaap Kamps, University of Amsterdam, The Netherlands Scientific Committee Florian Boudin - Nantes University, France Lynne Bowker - University of Ottawa, Canada Sara Carvalho - Universidade NOVA de Lisboa / Universidade de Aveiro, Portugal Rute Costa - Universidade NOVA de Lisboa, Portugal Eric Gaussier - University Grenoble Alpes, France Natalia Grabar - CNRS, France Ana Ostroški Anić - Institute of Croatian Language and Linguistics, Croatia Tatiana Passali - Aristotle University of Thessaloniki Grigorios Tsoumakas - Aristotle University of Thessaloniki Sara Vecchiato - University of Udine, Italy Cornelia Wermuth - KU Leuven, Belgium ##################### Contact ##################### For inquiries, please contact: giorgiomaria.dinunzio@unipd.it |
|
