Analyzing Behavior Data of Students in the Classroom with Association Rule Mining

Students behavior in the classroom has a close relation with the academic performance of students. With the widespread use of mobile phones, the students are easily distracted by mobile phones during the class. Therefore, analyzing the students behavior in the classroom becomes important to improve the quality of teaching. However, the existing methods for analyzing students behavior in the classroom, such as facial recognition, has low efficiency with large amounts of irrelevant data. And it also raises concerns about privacy and ethical issues. To solve this problem, in this paper, a novel mobile software called EasyClass has been developed to collect and quantify students behaviors. Based on the data collected by EasyClass, we propose to utilize user profiling to study students behavior in the classroom. The clustering algorithm is firstly employed to classify the data and generate semantically tags which describe students behavior in the classroom. Then a sequence and association rule analysis method is used to analyze the behavior tags and generated association rules. The results enable educators and non-expert users to gain a deeper understanding of students behavior in the classroom, and further improve the quality of the classroom teaching.