Article: Generative AI and multimodal learning spaces for perceiving and regulating English learning anxiety Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.65 pp.65 - 92 Abstract: This study has constructed a generative artificial intelligence-driven multimodal learning space for real-time perception and regulation of the anxiety states of English learners. By integrating facial expression analysis, speech feature extraction, and behaviour data modelling, the system achieved an accuracy rate of 87% in identifying learning anxiety, which was over 9% higher than that of single-modal methods. In a six-week intervention experiment, the state anxiety scores of the experimental group decreased by 31.6% compared to the control group, while the oral fluency of the experimental group improved by 24.3%. The research proves that the collaborative intervention of multimodal emotion computing and generative artificial intelligence can effectively break through the bottlenecks of delayed anxiety identification and single intervention methods in traditional teaching, providing a feasible path for the emotional adaptation of intelligent language learning environments. Inderscience Publishers - linking academia, business and industry through research

Title: Generative AI and multimodal learning spaces for perceiving and regulating English learning anxiety

Authors: Jian Zu

Addresses: School of Foreign Languages, Minnan University of Science and Technology, Quanzhou, 362000, China

Abstract: This study has constructed a generative artificial intelligence-driven multimodal learning space for real-time perception and regulation of the anxiety states of English learners. By integrating facial expression analysis, speech feature extraction, and behaviour data modelling, the system achieved an accuracy rate of 87% in identifying learning anxiety, which was over 9% higher than that of single-modal methods. In a six-week intervention experiment, the state anxiety scores of the experimental group decreased by 31.6% compared to the control group, while the oral fluency of the experimental group improved by 24.3%. The research proves that the collaborative intervention of multimodal emotion computing and generative artificial intelligence can effectively break through the bottlenecks of delayed anxiety identification and single intervention methods in traditional teaching, providing a feasible path for the emotional adaptation of intelligent language learning environments.

Keywords: generative artificial intelligence; multimodal learning; English learning anxiety; affective computing.

DOI: 10.1504/IJICT.2026.154183

International Journal of Information and Communication Technology, 2026 Vol.27 No.65, pp.65 - 92

Received: 05 Mar 2026
Accepted: 10 Apr 2026
Published online: 15 Jun 2026 *

Title: Generative AI and multimodal learning spaces for perceiving and regulating English learning anxiety

Keep up-to-date