Title: Detection method of students' classroom learning behaviour based on parallel classification algorithm

Authors: Degang Lai; Ke Wang

Addresses: Online and Continuing Education College, Southwest University, Chongqing, 400715, China ' Online and Continuing Education College, Southwest University, Chongqing, 400715, China

Abstract: In order to overcome the problem that students' learning behaviour process is easy to form misclassification in the process of serial classification, this paper proposes a method to detect students' learning behaviour in class based on parallel classification algorithm. The parallel classification model is constructed. By measuring Kinect coverage and adjusting Kinect top view angle, the coordinates of each student's position are transformed. The auxiliary feature vector is applied in behaviour recognition to realise the parallel combination and processing of multiple data sources, accurately extract the feature vector to form different relevance, and realise the detection of students' classroom learning behaviour. The experimental results show that students can grasp the degree of interest in the course and the degree of seriousness in the whole teaching process. The detection rate is more than 90%, which is practical.

Keywords: classroom learning behaviour; parallel classification algorithm; Kinect; skeleton feature vector.

DOI: 10.1504/IJCEELL.2022.10038646

International Journal of Continuing Engineering Education and Life-Long Learning, 2022 Vol.32 No.3, pp.279 - 294

Received: 21 Jan 2020
Accepted: 05 Aug 2020

Published online: 11 Jul 2022 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article