Research on machine analysis of classroom teacher-student interaction behavior
LIAO Shengbin1， QI Fei2
(1.National Engineering Laboratory for Educational Big Data， Central China Normal University， Wuhan 430079， China；2.National Engineering Research Center for E-Learning， Central China Normal University， Wuhan 430079， China)
Abstract：The main position of quality-oriented education is the classroom， with teachers and students as the main body. The most direct way for them to communicate is through classroom teacher-student interaction. Therefore， the interactive analysis between teachers and students has been highly concerned by educators. This study adopts the theory of improved Flanders interact analysis system (IFIAS) to code 23 classroom videos according to the coding rules. After the coding category statistics， principal component analysis and density clustering algorithm were used to realize the classification of classroom teacher-student interaction styles based on machine learning. And the classrooms were finally classified into three categories： student-centered， teacher-student interactive， and teacher-centered. This study not only extends the theoretical value of the Flanders interaction analysis system but also provides a new way of analysis for classroom analysis. Many classroom related features were generated through this analysis， which would improve the efficiency of classroom teacher-student interaction analysis and provide teachers with directions for improving classroom teaching.
2024, Vol. 58 
