UniversalCEFR: Enabling Open Multilingual Research on Language Proficiency Assessment
University of Bath · Cardiff University and King Abdul Aziz University · Universität Bielefeld · University of Exeter · Gothenburg University · Cardiff University · Mohamed bin Zayed University of Artificial Intelligence · Brigham Young University · ISCTE - Instituto Universitário de Lisboa and INESC-ID Lisboa · Universitat Pompeu Fabra and Universitat Pompeu Fabra · Göteborg University · National Research Council Canada · UCL
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。DOI 10.18653/v1/2025.emnlp-main.491 ↗
摘要
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.