직업훈련 공급의 구조적 불균형과 시장 신호분석: 행정 데이터 및 웹 크롤링 분석을 중심으로

초록

This study examines the structural characteristics of Korea’s vocational training supply by analyzing 2,091,159 participants recorded in the 2025 administrative dataset from the Korea Employment Information Service. Unlike prior research that has predominantly focused on demand-side outcomes, this study adopts a supply-centered analytical framework. It investigates temporal, regional, and sectoral patterns using large-scale administrative data, web crawling, and text mining. The temporal distribution reveals a pronounced concentration in April–May, which together account for 24.3% of the annual supply, while December exhibits the lowest share at 4.9%, suggesting a monthly rather than a period of concentration. Despite these fluctuations, overall annual dispersion remains relatively moderate, as indicated by an HHI of 0.0899 and a Gini coefficient of 0.142. Regionally, the Seoul metropolitan area accounts for 66.8% of the total supply, with per-capita availability nearly twice that of non-metropolitan regions, highlighting a substantial spatial imbalance. The share of digital and emerging technology fields increases markedly from 11.2% in 2023 to 24.7% in 2025. Text mining identifies high-frequency keywords such as AI, data, cloud, and UX, reflecting strong market signaling. These findings underscore the need for temporally balanced provision, regional infrastructure enhancement, and sustained investment in digital skill development to ensure a more equitable and responsive training supply system.

키워드