Student Modeling for Adaptive Agent to Enhance Interest and Comprehension
Sung‐il Kim, Myung-Jin Lee, Woogul Lee, Karam Lim, Cheon‐woo Han, YeonHee So, Sunyoung Lee, Su-Young Hwang, Sung-Hyun Yun, Dong-Seong Choi · Proceedings of the Annual Meeting of the Cognitive Science Society · 2001
Student Modeling for Adaptive Agent to Enhance Interest and Comprehension Sung-il Kim * ([email protected]), Myung-Jin Lee * ([email protected]), Woo-Gul Lee * ([email protected]), Karam Lim * ([email protected]), Cheon-woo Han * ([email protected]), Yeonhee So * ([email protected]), Sunyoung Lee * ([email protected]), Su-Young Hwang * ([email protected]), Sung-Hyun Yun ** ([email protected]), Dong-Seong Choi ** ([email protected]) Department of Education, Korea University, Seoul, 136-701, Korea Department of Information and Communication Engineering, Cheonan, 330-704, Korea Introduction The recent development of teachable agent provides the learner with an active role as a knowledge constructor and focuses on the individualization. Individualized agent provides differential interface and responses adaptively depending on the characteristics of user and its behaviors. The aim of the ‘adaptive agent’ is not only to maximize the learner’s cognitive functions but also to enhance the interests and motivation to learn. To improve adaptiveness and individualization of the agent, it is necessary to assess each user’s specific cognitive and motivational characteristics and ongoing response patterns during learning. In order to establish the Figure 1: Structure of nodes (d: duration, f: frequency, n: number) relationships among user characteristics and response patterns and to extract the algorithm among Conclusion variables, we measured the individual characteristics and The results suggest that certain type of learner responses or analyzed logs of the teachable agent named KORI (KORea the combination of the responses would be useful indices to university Intelligent agent) through the student modeling. predict the learners’ individual characteristics and ongoing learning outcome. This study proposed a new type of Method dynamic assessment for individual differences and ongoing cognitive/motivational learning outcomes through the Participants. computation of responses without measuring them directly. Twelve fifth graders (8 males and 4 females) participated in The individualization is the ultimate goal of developing the the student modeling. learning agent. Thus construction of individualized student model based on the ongoing response pattern of the user that Procedure. are highly correlated with the individual differences and All of the participants took the 30 minutes lesson on the learning outcome may be the useful methodology to ‘Rock Cycle’ which is about the three kinds of rocks and their understand the learner’s dynamic change during learning. transformation. Next, participants filled in questionnaires on individual characteristics including self-efficacy, goal Acknowledgments orientations, and metacognition. The final step is the 30 - 40 This research was supported by Brain informatics Research minute interaction with KORI. During this period, the log Program sponsored by Korean Ministry of Science and data were recorded automatically and participant’s behaviors Technology. were videotaped. After teaching KORI, participants completed the interest questionnaire and comprehension test. References Results Biswas, G., Schwartz, D., Bransford, J. & TAG-V. (2001). Technology support for complex problem solving: From SAD environment to AI. In Forbus and Feltovich, (Eds.), Smart machines in education, Menlo Park, CA: AAAI Press. Chi, M.T.H., Siler, S. A., Jeong, H., Yamauchi, T., & Hausmann, R. G. (2001). Learning from human tutoring. Cognitive Science, 25(4), 471-533. A correlation analysis was conducted to identify the relationships among individual characteristics, user responses, and learning outcomes. Among hundreds of possible relationships between numerous variables in three dimensions, nine key user responses were extracted, which were highly correlated with either individual characteristics and learning outcomes (See Figure 1).