Title: Student's classroom behaviour recognition method based on abstract hidden Markov model

Authors: Guojuan Li

Addresses: Department of Student Work and Social Sciences, Shijiazhuang University of Applied Technology, Shijiazhuang 050081, China

Abstract: In order to improve the standardisation of mutual information index, accuracy rate and recall rate of student classroom behaviour recognition method, this paper proposes a student's classroom behaviour recognition method based on abstract hidden Markov model (HMM). After cleaning the students' classroom behaviour data, improve the data quality through interpolation and standardisation, and then divide the types of students' classroom behaviour. Then, in support vector machine, abstract HMM is used to calculate the output probability density of support vector machine. Finally, according to the characteristic interval of classroom behaviour, we can judge the category of behaviour characteristics. The experiment shows that normalised mutual information (NMI) index of this method is closer to one, and the maximum AUC-PR index can reach 0.82, which shows that this method can identify students' classroom behaviour more effectively and reliably.

Keywords: classroom behaviour; hidden Markov model; HMM; abstract space; behaviour recognition; probability density; behaviour category.

DOI: 10.1504/IJITM.2024.139570

International Journal of Information Technology and Management, 2024 Vol.23 No.3/4, pp.232 - 243

Received: 29 Nov 2022
Accepted: 27 Feb 2023
Published online: 04 Jul 2024 *

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