nurlanm/UniSkill_dataset

UniSkill is a dataset for matching university curricula to professional competencies. It contains manually annotated and synthetic data linking university course content (sentences, title) with skills from the European Skills, Competences, Qualifications and Occupations (ESCO) taxonomy.

The original annotation guidelines used in the research are also provided in this repository.

For details about the dataset construction, annotation process, experiments, and evaluation, see the associated publication:

UniSkill: A Dataset for Matching University Curricula to Professional Competencies https://aclanthology.org/2026.lrec-1.31/

Citation:

@inproceedings{musazade-etal-2026-uniskill, title = "{U}ni{S}kill: A Dataset for Matching University Curricula to Professional Competencies", author = "Musazade, Nurlan and Mezei, J{'o}zsef and Zhang, Mike", editor = "Piperidis, Stelios and Bel, N{'u}ria and van den Heuvel, Henk and Ide, Nancy and Krek, Simon and Toral, Antonio", booktitle = "Proceedings of the Fifteenth Language Resources and Evaluation Conference", month = may, year = "2026", address = "Palma de Mallorca, Spain", publisher = "ELRA Language Resource Association", url = "https://aclanthology.org/2026.lrec-1.31/", doi = "10.63317/2n39qzvk2eqe", pages = "456--469", abstract = "Skill extraction and recommendation systems have been studied from recruiter, applicant, and education perspectives. While AI applications in job advertisements have received broad attention, deficiencies in the instructed skills side remain a challenge. In this work, we address the scarcity of publicly available datasets by releasing both manually annotated and synthetic datasets of skills from the European Skills, Competences, Qualifications and Occupations (ESCO) taxonomy and university course pairs and publishing corresponding annotation guidelines. Specifically, we match graduate-level university courses with skills from the Systems Analysts and Management and Organization Analyst ESCO occupation groups at two granularities: course title with a skill, and course sentence with a skill. We train language models on this dataset to serve as a baseline for retrieval and recommendation systems for course-to-skill and skill-to-course matching. We evaluate the models on a portion of the annotated data. Our BERT model achieves 87{%} F1-score, showing that course and skill matching is a feasible task." }