Author(s)

Yiqi Bai

Corresponding Author

Yiqi Bai

Abstract

Academic stress has become a significant concern for students worldwide, impacting their psychological well-being and academic performance. Traditional interventions such as counseling and therapy often face limitations in accessibility and scalability. This study explores the potential of AI large models, specifically GPT-based systems, in assisting students with emotional regulation to alleviate academic stress. The research introduces an AI-driven intervention tool, "EmoBuddy AI," designed to monitor and regulate emotions through personalized feedback. The tool integrates cognitive-behavioral techniques, emotional regulation strategies, and mindfulness exercises to support students in managing stress. A total of 200 students in Beijing participated in a 4-week experimental study, with 100 students in the experimental group using EmoBuddy AI, and 100 students in the control group receiving traditional counseling. The study evaluated the effectiveness of the AI intervention using the Perceived Stress Scale (PSS), the Emotional Regulation Questionnaire (ERQ), and physiological measures. Results showed a significant reduction in stress levels and improved emotional regulation among the experimental group, with PSS and ERQ scores indicating better outcomes compared to the control group. Physiological measures also revealed lower stress markers in the experimental group. This study highlights the potential of AI-based interventions in emotional regulation for students, providing a scalable and personalized solution for managing academic stress. The findings suggest that integrating AI technologies with socio-psychological frameworks can enhance emotional well-being and contribute to better academic outcomes.

Keywords

Cognitive-Behavioral Therapy (CBT), Academic Stress, Emotional Regulation, AI-Driven Intervention