Automated essay evaluator

• Main Authors: Escutin, Mabel Trinidad M., Estioko, Allan Patrick D., Plaza, Megann Ruth V.
• Format: text
• Language: English
• Published: Animo Repository 2004
• Subjects: Natural language processing (Computer science); Artificial intelligence; Information retrieval; Essay > Evaluation > Computer programs
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/8516
• Institution: De La Salle University

Evaluating essay portions of exams has always been labor-intensive for the teaching staff and human evaluators. To solve this problem, researches regarding automatic essay evaluation are ongoing, and practical implementation of such has risen just recently. This research presents Automated Essay Evaluator, a system that automates the evaluation of a large collection of essay-type documents using natural language processing (NLP) and Latent Semantic Analysis (LSA) technique. Rule-based natural language parsing is used for the grammar checking of the text while LSA is used to evaluate the content. Testing and evaluation were done to determine its performance in evaluating essays. Results show that the system is 85%-93% accurate in predicting the grammar grade, while it is 89%-98% accurate in predicting the content grade of essays.