Thematic Structure of Pedagogical Agent Studies: LDA Analysis for the Period 2020–2025

TURGUT Y. E., Aktı Aslan S., Kopuz T., Aslan A., Allison J., ÖZYURT Ö.

International Journal of Human-Computer Interaction, 2026 (SCI-Expanded, SSCI, Scopus) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Basım Tarihi: 2026
  • Doi Numarası: 10.1080/10447318.2026.2655931
  • Dergi Adı: International Journal of Human-Computer Interaction
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Social Sciences Citation Index (SSCI), Scopus, ABI/INFORM, Compendex, INSPEC, Psycinfo
  • Anahtar Kelimeler: bibliometric analysis, latent Dirichlet allocation, LDA, Pedagogical agents, topic modeling
  • Karadeniz Teknik Üniversitesi Adresli: Evet

Özet

This study examines 2,360 pedagogical agent (PA) articles published between 2020 and 2025 to identify thematic structures, trends, and research gaps through bibliometric analysis and LDA-based topic modeling. The optimal model produced 10 themes (c_v = 0.4767). The dominant themes—Learner Interaction with Virtual Assistants in Education, Student-Centred Learning and Individual Guidance, and NLP Foundations for Pedagogical Agents—account for 51.3% of publications. Trend analyses revealed consistent growth in both Artificial Intelligence and Pedagogical Design in Education and NLP Foundations for Pedagogical Agents, while Learner Interaction with Virtual Assistants in Education reached saturation. Interaction network results showed these themes at the conceptual core of the field, whereas AI-Driven Medical Chatbots, Knowledge-Based Learning Models, and Student Experiences in Intelligent Learning Environments emerged as future research areas. Overall, findings highlight a strong thematic shift toward artificial intelligence, natural language processing, and student-centered learning.