A personalized system of instruction is one of the strategies to personalize instruction. It is a technique that allows the student to move from one unit to another according to his own pace and his potential. Although this system is distinguished with activity and effectiveness to master the instructional subject, it lacks evaluation of the essay questions automatically. Automated essay scoring is the operation of scoring written essays by computer programs. It has been widely used in recent years. In this paper, a proposed method is presented to automatically grade students’ Arabic essays to support personalized systems of instruction. It uses the elitist-Jaya (e-Jaya) optimization algorithm to train the classic artificial neural network (called eJaya-NN). The proposed method is tested over 240 student’s essays. The essays are graded by two human experts in the fields then they are fed to a pre-processing phase to be converted to a digit’s matrix. The results are evaluated using different measures and it is compared with some optimization algorithms. The eJaya-NN outperformed all compared algorithms and achieved the best values. Its correlation with the scores of the human experts equals 0.92 which indicates that the proposed method produces acceptable scores for the Arabic essay compared to the human experts and can effectively increase the features of personalized systems of instruction.

个性化的教学系统是个性化教学的策略之一。这项技术可以让学生根据自己的步调和潜力从一个单元移动到另一个单元。尽管该系统在掌握教学主题方面具有活动性和有效性，但它缺乏自动评估论文问题的能力。自动论文评分是通过计算机程序对书面论文进行评分的操作。近年来已被广泛使用。在本文中，提出了一种建议的方法来自动对学生的阿拉伯文文章进行评分，以支持个性化的教学系统。它使用elitist-Jaya（e-Jaya）优化算法来训练经典的人工神经网络（称为eJaya-NN）。该方法在240篇学生的论文中得到了验证。论文由两位领域的专家进行评分，然后将其送入预处理阶段，然后转换为数字矩阵。使用不同的方法评估结果，并将其与一些优化算法进行比较。eJaya-NN的性能优于所有比较算法，并获得了最佳价值。它与人类专家的分数的相关性是0.92，这表明，与人类专家相比，该方法为阿拉伯文的论文产生可接受的分数，并且可以有效地提高个性化教学系统的功能。eJaya-NN的性能优于所有比较算法，并获得了最佳价值。它与人类专家的分数的相关性是0.92，这表明，与人类专家相比，该方法为阿拉伯文的论文产生可接受的分数，并且可以有效地提高个性化教学系统的功能。eJaya-NN的性能优于所有比较算法，并获得了最佳价值。它与人类专家的分数的相关性是0.92，这表明，与人类专家相比，该方法为阿拉伯文的论文产生可接受的分数，并且可以有效地提高个性化教学系统的功能。