Essay Scoring for Essay Evaluation Using Support Vector Machine in predicting Youth Break the Boundaries Scholarship Applicants

Abstract

This study proposes an automated essay scoring system utilizing Support Vector Machine (SVM) to predict successful candidates for the Youth Break the Boundaries Scholarship. The system aims to address the time-consuming and subjective nature of evaluating scholarship applicants' essays. A dataset comprising essays from previous applicants and corresponding human evaluation scores is utilized to train the SVM model. The model incorporates various textual features such as essay length, grammar, vocabulary, coherence, and argument structure. Performance evaluation is conducted through cross-validation techniques and compared against human assessment scores. The results demonstrate that the SVM-based automated essay scoring system achieves a high degree of accuracy in predicting scholarship applicants who meet the evaluation criteria. By leveraging machine learning techniques, this approach reduces the burden of manual evaluation, ensures consistent and objective scoring, and accelerates the selection process. Furthermore, the study identifies the most influential features in determining essay quality and applicant suitability. These insights can assist scholarship committees in refining their evaluation criteria and improving the selection process. The proposed automated essay scoring system enhances efficiency while maintaining fairness and objectivity in the scholarship evaluation process. It allows committees to evaluate a larger pool of applicants accurately and efficiently, providing deserving candidates with a fair opportunity to receive the Youth Break the Boundaries Scholarship.