Learning Attitude Analysis based on Operation Logs of Fill-in Workbook System

The authors are developing a fill-in workbook system that allows faculty members to ascertain the attitude of all students to classes including those who are not active, and improve lectures through well-timed and appropriate actions. To support the actions of teachers to improve lessons and teaching materials, we analyzed the operation logs of students using our Fill-in Workbook System in multiple lessons. The transition of learning attitudes for each lesson and for each time / each slide in the lesson for the entire class is extracted. In addition, to support the student's retrospective review of the lessons, the transition in the attitude of each student to the class is extracted. The results confirmed that lessons in which the class attitude became worse could be detected by analyzing the system focus time and the degree of synchronization between the page explained by the teacher and the page referenced by the student, and that the difference in learning attitudes increased when there were few blanks. In addition, we analyzed the degree of asynchrony (jumping ahead, falling behind) with the teacher's explanation and clarified the relationship between the time taken to explain one page and learning attitude. Furthermore, it was confirmed that the student's individual learning state could be divided into three separate conditions.