Salvato in:
| Autori principali: | , , , , , , , , , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2506.01419 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911156123205632 |
|---|---|
| author | Imperial, Joseph Marvin Barayan, Abdullah Stodden, Regina Wilkens, Rodrigo Sanchez, Ricardo Munoz Gao, Lingyun Torgbi, Melissa Knight, Dawn Forey, Gail Jablonkai, Reka R. Kochmar, Ekaterina Reynolds, Robert Ribeiro, Eugénio Saggion, Horacio Volodina, Elena Vajjala, Sowmya François, Thomas Alva-Manchego, Fernando Madabushi, Harish Tayyar |
| author_facet | Imperial, Joseph Marvin Barayan, Abdullah Stodden, Regina Wilkens, Rodrigo Sanchez, Ricardo Munoz Gao, Lingyun Torgbi, Melissa Knight, Dawn Forey, Gail Jablonkai, Reka R. Kochmar, Ekaterina Reynolds, Robert Ribeiro, Eugénio Saggion, Horacio Volodina, Elena Vajjala, Sowmya François, Thomas Alva-Manchego, Fernando Madabushi, Harish Tayyar |
| contents | We introduce UniversalCEFR, a large-scale multilingual and multidimensional dataset of texts annotated with CEFR (Common European Framework of Reference) levels in 13 languages. To enable open research in 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 modelling across tasks and languages. To demonstrate its utility, we conduct benchmarking 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 support using linguistic features and fine-tuning pretrained models in multilingual CEFR level assessment. Overall, UniversalCEFR aims to establish best practices in data distribution for language proficiency research by standardising dataset formats, and promoting their accessibility to the global research community. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_01419 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | UniversalCEFR: Enabling Open Multilingual Research on Language Proficiency Assessment Imperial, Joseph Marvin Barayan, Abdullah Stodden, Regina Wilkens, Rodrigo Sanchez, Ricardo Munoz Gao, Lingyun Torgbi, Melissa Knight, Dawn Forey, Gail Jablonkai, Reka R. Kochmar, Ekaterina Reynolds, Robert Ribeiro, Eugénio Saggion, Horacio Volodina, Elena Vajjala, Sowmya François, Thomas Alva-Manchego, Fernando Madabushi, Harish Tayyar Computation and Language We introduce UniversalCEFR, a large-scale multilingual and multidimensional dataset of texts annotated with CEFR (Common European Framework of Reference) levels in 13 languages. To enable open research in 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 modelling across tasks and languages. To demonstrate its utility, we conduct benchmarking 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 support using linguistic features and fine-tuning pretrained models in multilingual CEFR level assessment. Overall, UniversalCEFR aims to establish best practices in data distribution for language proficiency research by standardising dataset formats, and promoting their accessibility to the global research community. |
| title | UniversalCEFR: Enabling Open Multilingual Research on Language Proficiency Assessment |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2506.01419 |