Mining Online Homework Clickstream: Detecting Students' Cooperative Behavior

An algorithm model is proposed in this paper to detect learners' cooperative behaviors via online homework click actions, in which students’ clickstream sequence is turned into behavior sets. This paper proposes an event-based algorithm model to capture characteristic actions in student behavior sequences, including doing, checking and seeking. These characteristic actions were found of being closely associated with cooperative behaviors among students. A recognition algorithm model based on disjointed set is proposed in this paper, which features both practicality and efficiency in case of limited training data.