Analysis and extraction of behaviors by students in lectures.
Watanabe Eiji, Takashi Ozeki, Takeshi Kohama · Educational Data Mining · 2014
In this paper, we discuss the influence the following behaviors on the behavior by a specific student in lectures; (i) the behavior by the lecturer, (ii) the behaviors by other students, and (iii) the behavior by oneself. First, we detect features for behaviors by lecturer and students by using image processing methods. Next, the relations among the above features are approximated by neural networks. Finally, we analyze the interaction between behaviors by lecturer and students based on the internal representations and show the synchronization between students.