채용 기관의 현장실습 대학생 직업 인성 핵심요인 분석: XAI 기반 분석방법을 중심으로

초록

This study aims to identify the key employability traits of university students who have completed internship programs, which are increasingly regarded as essential for entry-level employment in South Korea. The findings offer practical implications for both employers seeking effective recruitment strategies and young job seekers preparing for the labor market. While previous studies have primarily developed employability frameworks from an educational perspective, this study distinguishes itself by analyzing actual evaluation data collected by host organizations, based on the Ministry of Education’s standardized internship guidelines. The dataset consists of performance evaluations for 814 student interns from a university in Seoul. The data were categorized by job type—service-oriented and technical—and analyzed using hierarchical clustering and the XGBoost-SHAP algorithm, a widely used technique within explainable artificial intelligence (XAI). This approach enabled the identification of the relative importance of each employability trait. The results reveal that `organizational adaptability` and `proactiveness` were the most important traits in service roles, whereas `diligence` and `analytical judgment` were key in technical roles. `Interpersonal skills` were found to be consistently important across both categories. In contrast, traits such as `creative planning`, `academic performance (GPA)` and `prior social experience` were found to be of relatively low significance. By applying a data-driven XAI approach to internship evaluations, this study contributes empirical insights into predicting both job fit and organizational fit. It also highlights the potential of explainable AI methods in bridging the gap between education and employment in future research.

키워드