고등교육에서 AI-교수자-학습자 상호작용 해외 연구 동향 분석: 토픽모델링을 중심으로

초록

This study systematically analyzed international research trends on AI-instructor-learner interactions in higher education. LDA topic modeling, time-series analysis, and co-occurrence network analysis were applied to 359 articles from major academic databases. Four research topics emerged: 'Instructor-AI collaborative instructional design and cognitive support,' 'Learner-AI direct interaction and response evaluation,' 'AI-mediated interaction in foreign language learning,' and 'AI-facilitated peer learning and socio-emotional engagement.' Time-series analysis showed that only foreign language learning research exhibited a significant upward trend, while learner-AI direct interaction stabilized after an initial surge. Co-occurrence network analysis showed learner-AI direct interaction as a central hub connecting other areas, while peer learning and socio-emotional engagement formed a relatively independent domain. Beyond early explorations of technology adoption, AI in education research has diversified across agents, contexts, and structures of interaction, with learner-AI direct interaction serving as a foundational link among research areas, and peer learning and socio-emotional engagement remaining relatively distinct. The study proposes expanding instructors' orchestrator roles, supporting quality and critical learner-AI interaction, refining discipline-specific interaction design, and broadening research on peer learning and socio-emotional engagement.

키워드