Application of Data Mining Techniques for Predicting Student Success in English Exit Exam

Abstract

The students' English proficiency has become an important requirement for job seeking after graduation. The universities in non-native English speaking countries find their challenges in improving their students' English language skills. Loei Rajabhat University has dealt with this issue for a long time by delivering various English language courses and tests to students. These activities have been carried out repeatedly year by year as a common routine. However, the real status of student success and the predictors of this issue have never known. This research, therefore, explored the available data: English test results (English placement test and exit exam) and data of students, who graduated in 2013, 2014, and 2015 by using the decision tree technique (C4.5). The research constructed and tested classification models for predicting student success in English exit exam. The research results also suggest that the English placement test result was a key attribute for predicting the result of English exit exam. These two attributes were plotted against each other by using the scatter plot where the regression analysis was carried out and the regression line and equation were generated to predict the students' English exit exam scores.