Preparing Modernisation Decisions of Public IT Recruitment through a Semi-Automated NLP Approach to Mapping Gov-ernmental Syllabi onto ESCO and e-CF

The rapid digitalisation of public services and the emergence of artificial intelligence, data science, and cybersecurity have exposed limitations in existing competency frameworks for public-sector IT professionals. In Spain, the long-standing examination syllabus for Level A IT civil servants is undergoing reform to address the growing misalignment between assessed competencies and current professional requirements. This paper presents a semi-automated, human-in-the-loop process to support decision-making by mapping the existing syllabus and authoritative European professional profiles to two complementary frameworks: the European Skills, Competences, Qualifications and Occupations (ESCO) taxonomy and the European e-Competence Framework (e-CF, EN 16234-1). The pipeline uses Sentence-BERT (S-BERT) with MPNet-based embeddings and cosine similarity to identify relevant skills and competences. Before automated matching, specialised profiles in artificial intelligence, data science, and cybersecurity, including references published by the European Union Agency for Cybersecurity (ENISA), are harmonised using consistent ESCO and e-CF terminology. The resulting candidate profiles are ranked for stakeholder consideration and consolidated through expert validation. By combining semantic matching, multi-source harmonisation, and expert judgement, the proposed approach reduces complex information to a manageable set of profiles, supporting decisions on candidate selection and competency development in public-sector IT. The methodology provides a replicable approach for modernising public recruitment in alignment with European competency frameworks.