Development of the Student Dependency Detection System―Using the Time-series Analysis of the Similarities of Reports

Abstract

It has been recognized that some students submit a report which has the high similarity compared with the other students. And many studies have been proposed to detect those reports, but these studies have focused on the similarity of one exercise. To detect the real dependency of student, it is necessary to analysis the transition of similarity in many exercises. In this study, we propose and develop the system which detect dependencies of student using time-series analysis of the similarities of reports.