Improved Student Modeling and Data Clustering for Personalized Online Teaching of International Chinese Teachers: The MKmeans Algorithm Approach

Abstract

This study explores personalized online teaching for international Chinese education by enhancing student modeling through advanced data mining. A novel MKmeans clustering algorithm, based on mean shift theory, is proposed to improve clustering stability and noise resilience. Experiments on datasets like Iris and Wine demonstrate MKmeans outperforms classical Kmeans in accuracy and F-measure. The findings enable adaptive teaching strategies, offering significant insights for optimizing online learning systems and advancing personalized education.