This paper reviews the literature in cognitive and learning science on collaborative learning focusing on the literature of Chi’s ICAP framework. The paper introduces how the framework has been defined to capture the deepening process of learning in collaborative learning and its implications towards developing learner support systems. This paper also reviews the literature of the authors ongoing work, using ICAP framework to (a) capture leaner’s state and (b)provide effective facilitation prompts to facilitate learning outcomes. This paper discusses how to design tutoring systems, developing education systems and conduct data analysis using ICAP framework to facilitate collaborative learning.

There has been increasing attention on inquiry learning, project-based learning, and design thinking as “knowledge creation” in various educational sectors in Japan, such as high school education, higher education, and professional education. However, learning as knowledge creation requires learners to not only participate in knowledge-creating activities but also to join in a knowledge-creating culture. Furthermore, knowledge building, which is a theory for learning as knowledge creation, has been used in research on classroom practices, learning environment design, technologies, and analysis methods over the past thirty years. These results contribute to discussions on how to support learners’ engagement in sustainable idea improvement. Consequently, this paper introduces knowledge-building research on computer-supported collaborative learning (CSCL) and recent learning analytics studies.

“Monsakun” is an intelligent learning support system for arithmetic word problems. Its effectiveness has been confirmed in continuous and long-term practical use. The system provides problem-posing assignments. Some cards are provided to the learners, who assemble an arithmetic word problem using the cards. When setting such assignments, teachers are required to analyze the arrangement of the cards into arithmetic word problems. However, this is difficult because teachers need to check all combinations of the cards. In this study, we developed a problem-posing simulator and a problem-posing assignment analyzer for Monsakun. The simulator can generate all combinations of the cards. The analyzer can identify all error patterns and correct patterns in the assignment. We also conducted an experiment to evaluate the analyzer by using past data of practical uses. The results suggest that the analyzer reveals features of the assignments that the teachers were unaware of.

In this paper, we propose a class design for a flip-flop learning using an adaptive learning support system. Our model consists of combination of flip-flop learning and active learning. Two policies are designated to promote learners’ autonomous learning through several-week lessons employed to master a single learning unit. The first policy mainly describes preparatory learning using the system that manages all learning materials covering three learning objectives, that is, knowledge acquisition, utilization, and application. Using this system, learners can learn in accordance with their understanding degrees. The second policy describes face-to-face learning including a check test performed at the first step of the class, a groupwork for knowledge sharing, and a reflection for the class activity. We evaluate the learning effectiveness of our proposed class design through a case study of programming classes. The results show that the class design contributes to improvement of learners’ knowledge and skills for programing.