An Automated Essay Scoring Based on Neural Networks to Predict and Classify Competence of Examinees in Community Academy

Abstract

AES has been widely used in assessing student learning outcomes. However, few studies use Automated Essay Scoring (AES) to simultaneously determine the community academy's competency test scores and levels. This study aims to apply AES to assess essays on the competency certification test. The AES can predict the examinees' scores and classify examinees' competency levels. The method used to build AES uses Back Propagation Neural Networks (BPNN). BPNN was chosen because of its simplicity and ease in building the model. The results showed that the AES for predicting the examinee's competency value showed the MAE value is 0.061621 and the accuracy value is = 97.9665 %. The results of the classification of student competency levels show Accuracy= 0.9063, Precision= 0.9167, Recall= 0.8888, and F1 Score= 0.8857.