디지털·AI 기반 영어교육 연구 동향: 텍스트 네트워크 분석 및 토픽모델링 중심으로

Abstract

This study aimed to discern trends in research related to digital and AI-based English education by examining studies published between 2011 and 2021. We analyzed 337 collected documents using a text-based network analysis method, specifically topic modeling. The findings include: 1) An examination of the year-by-year trend revealed a consistent focus on digital and AI-based English education since 2011, with a noticeable uptick from 2019 and a significant surge starting in 2020. 2) Words appearing with the highest frequency were ‘learning,’ ‘learners,’ ‘online,’ ‘college,’ ‘language,’ ‘program,’ ‘effect,’ ‘activity,’ ‘participant’, and ‘AI’, in that order. 3) Word-network analysis indicated that co-occurrence frequency was highest for word pairs such as ‘learninglanguage,’ ‘college-Korean,’ ‘learning-online,’ ‘learning-teaching,’ ‘learning-model,’ ‘video-lecture,’ and ‘teaching-method’. 4) LDA (Latent Dirichlet Allocation) topic modeling identified six primary themes: ‘instructional models’, ‘AI applications,’ ‘reading strategies and evaluation,’ ‘online course and content,’ ‘writing feedback,’ and ‘factors and effects of individualized learning’. The volume of papers suggests that the most researched topic was the factors and effects of individualized learning. The implications and suggestions derived from these findings are also discussed.