본문 바로가기
  • Home
Basic Search

A Study on Design Principles for Generative AI-Based Korean Language Classes

  • Korean Language & Literature
  • 2026, (133), pp.517~551
  • Publisher : Korean Language & Literature
  • Research Area : Humanities > Korean Language and Literature
  • Received : June 20, 2026
  • Accepted : July 20, 2026
  • Published : July 31, 2026

An, Sung-min 1

1한국과학기술원

Accredited

ABSTRACT

This study aims to derive design principles for generative AI-based Korean language classes based on the perceptions and needs of foreign learners of Korean. To this end, responses to four open-ended questions were analyzed from a survey of 150 foreign learners enrolled at a science and engineering university. The numbers of respondents were 107 for difficulties experienced in learning Korean, 95 for the types of learning support expected from generative AI, 75 for concerns about the use of AI, and 33 for additional opinions on AI-assisted Korean language classes. The responses were divided into meaning units and categorized, and responses containing more than one idea were coded into multiple categories. The results showed that vocabulary limitations and memorization were the most frequently reported difficulties in learning Korean, followed by grammar and usage, sentence construction and writing, speaking, pronunciation and listening, learning materials, review and time constraints, and natural expressions and cultural context. Regarding the support expected from generative AI, learners frequently mentioned vocabulary, translation and expression support, sentence correction and writing feedback, grammar explanations, and personalized learning materials and plans. The most prominent concern was overdependence on AI and the resulting reduction in learners’ thinking and practice. Other concerns included the scope and regulation of AI use, response accuracy, plagiarism and assignment ethics, reduced face-to-face interaction, and the naturalness and level appropriateness of AI-generated responses. In the additional comments, learners emphasized that AI should be used selectively according to specific purposes and situations, while also recognizing its potential for post-class review, personalized learning, and preparation for real-life situations. Based on these findings and theoretical discussions of scaffolding, self-regulated learning, AI feedback literacy, and teacher-guided technology integration, this study proposes seven design principles: prioritizing learner output, providing explanation-centered feedback, comparing and reviewing AI feedback, linking pre-class, in-class, and post-class learning, using AI for real-life situations, ensuring ethical use, and integrating AI through teacher guidance. Generative AI-based Korean language classes should be designed as a process that moves from learners’ initial production to AI-generated feedback and explanations, learner comparison and review, teacher feedback, and learner revision and reproduction. Within this structure, generative AI can function not as a substitute for learner performance, but as a learning resource that supports language production and revision, review, and real-life practice both inside and outside the classroom.

Statistics

  • 2 Viewed
  • 0 Downloaded

Tools

Issue List

Citation status

* References for papers published after 2025 are currently being built.

Basic Search Advanced Search