| Abstract |
In an era of rapidly evolving labor market demands, aligning policy documents with the skills required for emerging and existing occupations has become increasingly critical. This paper presents a scalable, AI-driven framework for skill mapping that integrates advanced sentence-embedding models, FAISS for high-speed similarity searches, and the European Skills/Competences, Qualifications and Occupations (ESCO) classification. By automatically extracting and analyzing skill references in policy texts, the framework helps policymakers and analysts identify recurring competencies, detect emerging themes (e.g., sustainability or digital literacy), and pinpoint potential workforce gaps. Additionally, it introduces a systematic method for assessing occupation-level relevance-calculating the overlap between policy-cited skills and ESCO-defined occupations to guide targeted upskilling and reskilling efforts. Empirical results suggest that this AI-enabled approach can markedly enhance both the speed and accuracy of policy analysis compared to traditional manual reviews, ultimately supporting data-driven decision-making at scale. |
