ScholarWorks@숙명여자대학교

초록

This study investigated strategies to secure sustainable employability for middle-aged workers amid labor-market changes driven by the proliferation of generative AI. Based on crystallized intelligence, the experiential asset possessed by middle-aged workers, it proposes a Human-Centric Skills (HCS)-based job redesign model that harmoniously connects human strengths with AI technology. For the research method, text mining was conducted on academic literature (RISS) and job market data from 2020–2022 (Period 1) and 2023–2025 (Period 2) to analyze changes in keyword frequency, competency hierarchies, and network structures. To ensure the qualitative validity of the analyzed data, job postings with detailed job descriptions were selected, and biases arising from sample size differences were statistically corrected using the TF-IDF weighting model. It was confirmed that academic discourse has shifted from a past welfare-centered orientation to a technology- and competency-centered one, and in the job market, human-centric skills such as communication, collaboration, and adaptability account for 45.2% of the overall required competencies. These results empirically confirmed that, unlike the 'Data' domain replaced by AI, the combination of accumulated experience ('Thing') and contextual coordination ('People') is the core pathway through which the crystallized intelligence of middle-aged workers is manifested. Additionally, the keyword network evolved from a fragmented form into a hyper-connected structure integrating AI and human-centric skills. However, a structural mismatch appeared: while academia emphasizes macro-level digital transformation, the job market demands concrete, field-oriented practical competencies such as facility management, administrative practice, and caregiving. Specifically, it was identified that while academic discourse focused on abstract 'adaptation,' the job market prioritizes skilled 'management' capabilities to complete practical tasks utilizing AI tools. To resolve this, this study proposed: first, the introduction of a ‘Data-HCS integrated PBL’ curriculum; second, job redesign based on an ‘AI-Augmented Manager’ model, where AI performs standardized tasks, and humans handle verification and coordination; and third, the establishment of an agile ‘hybrid skill dictionary’. Finally, to supplement the limitations of quantitative text analysis, this study suggested the need for future qualitative research through in-depth interviews and follow-up verification through the expansion of industry-specific data.

키워드