Applying Ensemble Classifier, K-Nearest Neighbor and Decision Tree for Predicting Oral Reading Rate Levels

Abstract

For many years, reading rate as word correct per minute (WCPM) has been investigated by many researchers as an indicator of learners’ level of oral reading speed, accuracy, and comprehension. The aim of the study is to predict the levels of WCPM using three machine learning algorithms which are Ensemble Classifier (EC), Decision Tree (DT), and K- Nearest Neighbor (KNN). The data of this study were collected from 100 Kurdish EFL students in the 2nd-year, English language department, at the University of Duhok in 2021. The outcomes showed that the ensemble classifier (EC) obtained the highest accuracy of testing results with a value of 94%. Also, EC recorded the highest precision, recall, and F1 scores with values of 0.92 for the three performance measures. The Receiver Operating Character curve (ROC curve) also got the highest results than other classification algorithms. Accordingly, it can be concluded that the ensemble classifier is the best and most accurate model for predicting reading rate (accuracy) WCPM.