| Download | - View final version: UniversalCEFR: enabling open multilingual research on language proficiency assessment (PDF, 1.7 MiB)
|
|---|
| DOI | Resolve DOI: https://doi.org/10.18653/v1/2025.emnlp-main.491 |
|---|
| Author | Search for: Imperial, Joseph Marvin; Search for: Barayan, Abdullah; Search for: Stodden, Regina; Search for: Wilkens, Rodrigo; Search for: Muñoz Sánchez, Ricardo; Search for: Gao, Lingyun; Search for: Torgbi, Melissa; Search for: Knight, Dawn; Search for: Forey, Gail; Search for: Jablonkai, Reka R.; Search for: Kochmar, Ekaterina; Search for: Reynolds, Robert Joshua; Search for: Ribeiro, Eugénio; Search for: Saggion, Horacio; Search for: Volodina, Elena; Search for: Vajjala, Sowmya1ORCID identifier: https://orcid.org/0000-0002-4033-9936; Search for: François, Thomas; Search for: Alva-Manchego, Fernando; Search for: Tayyar Madabushi, Harish |
|---|
| Affiliation | - National Research Council Canada. Digital Technologies
|
|---|
| Format | Text, Article |
|---|
| Conference | 2025 Conference on Empirical Methods in Natural Language Processing, November 4-9, 2025, Suzhou, China |
|---|
| Abstract | We introduce UniversalCEFR, a large-scale multilingual multidimensional dataset of texts annotated according to the CEFR (Common European Framework of Reference) scale in 13 languages. To enable open research in both automated readability and language proficiency assessment, UniversalCEFR comprises 505,807 CEFR-labeled texts curated from educational and learner-oriented resources, standardized into a unified data format to support consistent processing, analysis, and modeling across tasks and languages. To demonstrate its utility, we conduct benchmark experiments using three modelling paradigms: a) linguistic feature-based classification, b) fine-tuning pre-trained LLMs, and c) descriptor-based prompting of instruction-tuned LLMs. Our results further support using linguistic features and fine-tuning pretrained models in multilingual CEFR level assessment. Overall, UniversalCEFR aims to establish best practices in data distribution in language proficiency research by standardising dataset formats and promoting their accessibility to the global research community. |
|---|
| Publication date | 2025-11 |
|---|
| Publisher | Association for Computational Linguistics |
|---|
| Place | Stroudsburg, Pennsylvania, United States |
|---|
| Licence | |
|---|
| In | |
|---|
| Language | English |
|---|
| Peer reviewed | Yes |
|---|
| Export citation | Export as RIS |
|---|
| Report a correction | Report a correction (opens in a new tab) |
|---|
| Record identifier | 6f760d65-0398-413b-96b0-c1e3bf3539f2 |
|---|
| Record created | 2025-11-28 |
|---|
| Record modified | 2026-02-19 |
|---|
