English Composition Evaluation by KNN Algorithm Combined with Regression Modeling

Aiming at the time-consuming problem in the evaluation process of English composition, the study constructs an English composition evaluation model based on the K-Nearest Neighbors (KNN) algorithm and regression model. The study first improves the KNN algorithm using information entropy and uses the improved KNN as a classification study of evaluation data. Then the evaluation classification was integrated with the regression model and used to construct the English composition evaluation model. The results show that the evaluation data classification accuracy and recall of the English composition evaluation model are 91.57% and 92.86% respectively, and the highest precision of the evaluation model is 93.59%. This indicates that the English composition evaluation model constructed based on KNN and regression model can significantly improve the accuracy and reliability of the evaluation of English compositions, and can reduce the work pressure of English teachers to a certain extent. The study aims to provide an effective way for the development of English education, so as to improve the overall effect of English teaching.