This study aims to fill in the gap by exploring the interrelationship among academic emotions, classroom engagement and self-efficacy in EFL learning among Chinese non-English majors in smart classrooms. A total of 409 non-English majors in their first year from a Chinese university completed the revised version of the academic emotions scale, classroom engagement scale and self-efficacy scale and their interrelationship has been analysed by Structural Equation Modelling. Positive emotion significantly enhances classroom engagement, while negative emotion does not. Classroom engagement has a significant effect on promoting self-efficacy, which is influenced by both positive and negative emotions, with positive emotions having a greater impact. Specifically, classroom engagement mediates the relationship between positive emotion and self-efficacy, whereas it does not mediate between negative emotion and self-efficacy. Smart classrooms can stimulate students’ positive academic emotions, increase their engagement in learning so as to enhance their self-efficacy.

This study investigates the impact of classroom environment on students' psychological safety and learning motivation and further explores how psychological safety influences motivation and how motivation, in turn, affects students' willingness to communicate. These relationships are analyzed using a quantitative approach through Structural Equation Modeling (SEM) implemented in R. While previous studies have often examined these variables in isolation, this study adopts an integrated approach to construct a comprehensive theoretical model, including classroom environment, psychological safety, learning motivation, and communication willingness. Moreover, it also considers how differences such as gender and urban-rural background may shape the ways in which the classroom environment affects students, offering insights into how these influences may vary across student groups. The findings are expected to offer both theoretical guidance and practical implications for improving classroom teaching approaches and stimulating greater student participation and learning achievements.

This study addresses the problems in traditional English literature teaching methods for Chinese English majors, proposing a new teaching approach based on smart education concepts to enhance learning effectiveness. An evaluation of a semester-long reform in teaching methods is conducted using a quantitative methodology. The findings reveal significant differences in learning outcomes between the experimental and control classes, suggesting the new model's positive influence on student academic performances. Machine learning algorithms are also used to analyze student classroom behavior, indicating a significant increase in active engagement. In conclusion, students' learning outcomes, as well as engagement, can be substantially improved by integrating smart education techniques into English literature instruction.