Abstract

The Automated Essay Grading (AEG) used in universities, companies, and schools that is based on Artificial Intelligence, Machine Learning, and Natural language Processing has the capability to improve the grading system to overcome cost, time and teacher effort in correcting the student’s essay questions and papers. AEG system widespread use due its cost, accountability, standards, and technology which lead to be used and applied for multiple languages such as English, French, among others, but limited research has been conducted to automate the Arabic essay grading. Therefore, this paper introduced an Arabic AEG . In this paper, we proposed a model for Arabic essay grading based on F-score to extract features from student answer and model answers along with the use of Arabic WordNet system (AWS ), which is a useful knowledge-based tool for semantic similarity measures and text similarity algorithm. The use of AWS to find all related words from student answers to give the answer of student a score. Students do not oppress in the mark because he did not write the same model answer exactly which lead to the improvement of Arabic AEG system to match human grading. The proposed model evaluated using Arabic essay dataset was prepared for this model, and the result shows that our proposed model produces result matches human grading .

Key words: Automated essay grading, support vector machine, Arabic WordNet, Cosine similarity, . Natural language processing