EUVIMEDEUVIMED entdecken
Crossref

A Comparative Analysis of Research Trends in Generative AI Education in Korea and China Using LDA Topic Modeling: Focusing on KCI and CNKI

Hee-Jeong Seo, Gwan-Ji Park

Korean Comparative Education Society · 30.06.2026 · 36 (3) · 69-100

DOI: 10.20306/kces.2026.6.30.69 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Korean Comparative Education Society, online erschienen am 30.06.2026. Verfasst von Hee-Jeong Seo und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

[Purpose] This study compares research trends in generative AI education in Korea and China after the emergence of ChatGPT. It examines journal distribution, keyword frequency, and topic structures, and draws implications for the future direction of generative AI education research in Korea. [Methods] A total of 7,480 articles were analyzed, including 1,613 KCI articles and 5,867 CNKI articles. Keyword analysis and LDA topic modeling were conducted using a synonym dictionary and TF-IDF weighting. Based on coherence scores and interpretability, five topics were identified for KCI and eleven for CNKI. [Results] First, research on generative AI education increased rapidly in both countries after 2023. Second, Korea and China differed in the distribution of journals publishing this research. Third, Korean studies mainly featured keywords related to learner competencies and subject-based applications, such as AI literacy, writing, and prompt engineering. In contrast, Chinese studies placed relatively greater emphasis on keywords reflecting technological environments and education policy, including large language models, teaching and learning innovation, ideological and political education, and DeepSeek. Fourth, topic modeling showed that writing-related learning support accounted for the largest share in Korea, along with topics such as AI digital competency education and AI-based creativity education. In China, innovation in higher and vocational education accounted for the largest share, while human-AI collaborative education, ideological and political education, and educational applications of DeepSeek also emerged as distinct topics. [Conclusion] Korean research was characterized by its focus on classroom applications and learner competencies, whereas Chinese research placed relatively greater emphasis on teaching and learning innovation, education policy, and domestic AI technologies. These findings suggest that generative AI education research in Korea should extend its scope to the educational use of domestic LLMs and stronger connections with education policy and support systems.

Zitieren

APA

Seo, H., & Park, G. (2026). A Comparative Analysis of Research Trends in Generative AI Education in Korea and China Using LDA Topic Modeling: Focusing on KCI and CNKI. Korean Comparative Education Society, 36(3), 69-100. https://doi.org/10.20306/kces.2026.6.30.69

Vancouver

Seo H, Park G. A Comparative Analysis of Research Trends in Generative AI Education in Korea and China Using LDA Topic Modeling: Focusing on KCI and CNKI. Korean Comparative Education Society. 2026;36(3):69-100. doi:10.20306/kces.2026.6.30.69

BibTeX
@article{Seo2026,
  author = {Seo, Hee-Jeong and Park, Gwan-Ji},
  title = {A Comparative Analysis of Research Trends in Generative AI Education in Korea and China Using LDA Topic Modeling: Focusing on KCI and CNKI},
  journal = {Korean Comparative Education Society},
  year = {2026},
  volume = {36},
  number = {3},
  pages = {69--100},
  doi = {10.20306/kces.2026.6.30.69}
}
Kostenlos starten →