Modeling of the students scenario on a learning course

Abstract

Adjusting the content to each student is a major issue in e-Learning. From this viewpoint, a learning course as a series of content also must be adjusted according to the performance of the students. We propose a method that combines clustering and decision tree learning for constructing scenarios of the students' actions. The global statuses of the students are reflected to the clusters, and the local and sequential actions of the students are reflected to the decision trees. The results of e-Learning tests gathered from Japanese junior high school students was processed by our proposed method. We graded the clusters by adaptation to the trees, and selected a set of clusters as a scenario for the students. These scenarios have a possibility of aiding the adjustment, and revision of learning courses. © 2008 Springer-Verlag Berlin Heidelberg.