Design and Implementation of Teaching Quality Assessment System for Universities Based on Data Mining Algorithms

Abstract

A teaching quality assessment system for universities based on data mining algorithms utilizes advanced analytical techniques to evaluate and enhance the effectiveness of teaching practices. By leveraging data mining algorithms, such as clustering, classification, and association rule mining, this system can analyze various educational data sources, including student performance metrics, course evaluations, and instructor feedback. This paper introduces Generative Pre-trained Data Analytics (GPDA) as a novel approach for assessing teaching quality in educational settings. GPDA leverages advanced data analytics techniques to predict teaching quality metrics, classify teaching quality data, and forecast various aspects of teaching effectiveness. Through comprehensive analysis and evaluation, GPDA demonstrates its accuracy, reliability, and versatility in capturing the complexities of teaching quality across diverse educational contexts. Through comprehensive analysis and evaluation, GPDA achieves a Root Mean Squared Error (RMSE) ranging from 0.152 to 0.190 and R-squared (R²) values ranging from 0.698 to 0.805, showcasing its accuracy and reliability in predicting teaching quality metrics across diverse educational contexts. Furthermore, strong positive correlations between the synthetic data generated by GPDA and real teaching quality data, with correlation coefficient values ranging from 0.87 to 0.92, validate the fidelity of GPDA-generated synthetic data.