Article: A personalised knowledge tracking graph neural network driven by learning psychological motivation Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.43 pp.63 - 83 Abstract: Most of the existing knowledge tracking methods ignore the dynamic influence of psychological states on the learning process, resulting in limited accuracy of personalised prediction. To this end, this paper proposes a psychomotivation-driven personalised knowledge tracking graph neural network. By integrating motivational factors such as concentration and curiosity and constructing a student-knowledge heterogeneous graph, it can simulate the learning process more accurately. Experiments on the assistments2012 and ednet public datasets show that psychomotivation-driven personalised knowledge tracking graph neural network has an average improvement of 1.2% in the area under the roc curve metric and 2.1% in the prediction accuracy compared to the optimal baseline model, and the improvement is statistically significant. This study provides an effective approach for achieving fine-grained learning state assessment that integrates cognition and emotion. Inderscience Publishers - linking academia, business and industry through research

Title: A personalised knowledge tracking graph neural network driven by learning psychological motivation

Authors: Yumin Wang; Yanlong Cui

Addresses: College of Mechanical Engineering, Nanjing University of Industry Technology, Nanjing, 210023, China ' Nanjing Chengxu Communication Engineering Co., Ltd., Nanjing, 210023, China

Abstract: Most of the existing knowledge tracking methods ignore the dynamic influence of psychological states on the learning process, resulting in limited accuracy of personalised prediction. To this end, this paper proposes a psychomotivation-driven personalised knowledge tracking graph neural network. By integrating motivational factors such as concentration and curiosity and constructing a student-knowledge heterogeneous graph, it can simulate the learning process more accurately. Experiments on the assistments2012 and ednet public datasets show that psychomotivation-driven personalised knowledge tracking graph neural network has an average improvement of 1.2% in the area under the roc curve metric and 2.1% in the prediction accuracy compared to the optimal baseline model, and the improvement is statistically significant. This study provides an effective approach for achieving fine-grained learning state assessment that integrates cognition and emotion.

Keywords: knowledge tracking; psychological motivation; graph neural network; GNN; personalised learning.

DOI: 10.1504/IJICT.2026.153386

International Journal of Information and Communication Technology, 2026 Vol.27 No.43, pp.63 - 83

Received: 30 Dec 2025
Accepted: 29 Jan 2026
Published online: 06 May 2026 *

Title: A personalised knowledge tracking graph neural network driven by learning psychological motivation

Keep up-to-date