인공지능 챗봇의 교육적 활용 연구 동향 분석: 활동이론을 중심으로 = A Review of Research on Artificial Intelligence Chatbot in Education through the Lens of Activity Theory

다국어 초록 (Multilingual Abstract)

In this study, the research trend of educational use of artificial intelligence chatbots were systematically analyzed to identify current issues and explore future directions based on activity theory. 34 papers were analyzed by using the activity system model as an analysis framework. The results are as follows. Research on chatbots has been sharply increased from 2019 in both domestic and international. For the number of learners, above of college students were the most in both domestically and internationally. The most common sample size was 50 or less in domestic and more than 100 in foreign countries. Artificial intelligence chatbots were most often used as a purpose-based system, text-based, learning tools in both domestic and abroad. chatbots were developed for using in foreign countries, while existing chatbots were used in Korea. In both domestic and foreign countries, chatbots were aimed to teach English and intellectual skill. The most result of learning is cognitive and affective domain in Korea and affective domain in foreign conturies. The most number of use was only once in Korea and twice in foreign countries. In both domestic and foreign countries, it was the most common case that the instructor were in control of environment and offered scaffolding while using chatbots. As for the learning environment, offline was the most common in Korea, and both online and offline were identically common in foreign countries. Chatbots were not used in cooperative learning at all abroad and only one case in Korea. Based on the results, the direction of future research is presented as follows. First, a follow-up study is required to develop a task-oriented, purpose-built chatbot and strictly verify its effectiveness. Second, it is necessary to develop chatbots that support the affective domain as agents. Third, it needs to design the cooperative learning in which chatbots function as tutors and provide timely feedback. Finally, research on learning design that can effectively use chatbots by utilizing learning analysis data on online learning environment is required.