Abstract: Automated Arabic Essay Scoring (AAES) systems have the potential to revolutionize educational assessment in the Arabic-speaking world, but their accuracy is often hindered by the complexities of Arabic morphology and semantics. This study introduces a novel AAES model, the Arabic Stemming Enhancement (ASE) algorithm, which integrates Arabic WordNet (AWN) to enhance semantic understanding and root extraction accuracy. By leveraging AWN's lexical database, ASE overcomes limitations of traditional stemming approaches, such as misinterpreting polysemous words. Rigorous evaluation on student essays and a WordNet-based dataset demonstrates ASE's superior performance compared to existing algorithms. Notably, ASE achieved a 0.977 Pearson correlation with human-assigned scores and a 15% increase in accuracy, alongside improvements in precision, recall, and F1-score. This research significantly advances Arabic Natural Language Processing (NLP), offering a more accurate, equitable, and scalable automated assessment tool for Arabic language education.
Keywords: Arabic Stemming Enhancement, Automated Essay Scoring, Arabic WordNet, Natural Language Processing, Semantic Analysis, Root Extraction, Educational Assessment, Machine Learning.
Title: A NOVEL ARABIC STEMMING ALGORITHM WITH WORDNET INTEGRATION FOR IMPROVED ESSAY SCORING
Author: Mohamad Alobed, Yazeed Al Moaiad
International Journal of Computer Science and Information Technology Research
ISSN 2348-1196 (print), ISSN 2348-120X (online)
Vol. 12, Issue 2, April 2024 - June 2024
Page No: 40-50
Research Publish Journals
Website: www.researchpublish.com
Published Date: 24-June-2024
