Abstract English is one of the essential contents in the higher education system, and the innovative research of English teaching has become a research hotspot in the academic world. This paper introduces the neural network model to the research of English classrooms in colleges and universities and constructs a neural network prediction model based on teacher-student interaction behavior. It is used to identify, classify, and predict the behavior of teacher-student interactions. Further, explores the impact of teacher-student interaction on student behavior and its quality on college English teaching. The neural network model in this paper achieves optimal performance in the long-term prediction of teacher-student interaction behavior, and the overall short-term prediction accuracy significantly exceeds that of other prediction models. The differences in grade level, gender, and number of lecturers were all significant for the quality of English classroom teaching in colleges and universities. There was a significant correlation between teacher-student interaction behaviors (classroom cooperation, classroom games, classroom communication, and situational interpretation) and English teaching quality (English achievement, classroom participation, and learning interest). Teacher-student interaction behaviors significantly improve English classroom teaching quality and predict teaching quality by up to 45.24% of the explained variance. Classroom cooperation is the primary factor in positive classroom interaction behavior between teachers and students.
Read full abstract