본문 바로가기
  • Home

Discovering Youth Job Search Bundles Using Association Rule Mining

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2025, 30(11), pp.319~327
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : September 8, 2025
  • Accepted : October 22, 2025
  • Published : November 28, 2025

Hyoungrae Kim 1 Jongdeok Lim 1 Jeongrae Kim 2

1한국고용정보원
2한국폴리텍대학 춘천캠퍼스

Accredited

ABSTRACT

Given the persistent challenges of youth unemployment and skills mismatch, a granular understanding of how young job seekers actually prepare for the labor market is critical for policy intervention. While most research on this topic examines the effectiveness of individual activities, understanding the combinations of these activities as strategic 'patterns' is crucial for developing effective employment policies. For this purpose, this study utilizes youth panel data, employing association rule mining (ARM) and logistic regression. The analysis reveals that job search patterns are distinctly differentiated by sociodemographic characteristics such as gender, educational attainment, and region. For instance, a strong association was found between {female residents in the metropolitan area} and {preparing for official English tests} (Lift=1.282). Furthermore, residents in the metropolitan area tended to focus on preparing for practical, job-related certifications. This suggests that youth job search activities are not uniform but are the result of strategic choices based on an individual's socio-structural position. These findings highlight the need for customized support policies tailored to the specific characteristics of different groups.

Citation status

* References for papers published after 2024 are currently being built.