Enhancing Grade Prediction in WE Learn Mobile Classroom Using Clustering-based Decision Tree

With the advent of the big data era, various industries are facing the trend of informatization, and the education industry is no exception. Traditional teaching methods are no longer the only option, and various educational platforms are emerging, leading to the continuous increase of data resources in the education industry. However, the value of these massive amounts of data has not been fully developed due to some restrictions in early data mining technology and personal data protection. Currently, most educational platforms can only store and query data simply, and these undeveloped data also lead to some educational platforms lacking complete functions in personal services. To solve this problem, this paper proposes a decision tree model improved by using clustering algorithms. Firstly, we conducted feature engineering and used clustering algorithms to divide students’ English scores. Next, we brought these feature data into the model, verified the influencing factors of English scores, constructed a decision tree model, and achieved the prediction of WE Learn’s students’ course scores. Finally, we comprehensively evaluated the model using common model evaluation indicators, MAE and RMSE. The results show that the constructed model can provide some reference value for the English education and teaching work of art students on the WE Learn platform.