A Comprehensive Survey on Methodologies for Automatic Essay Scoring

In the field of educational endeavors, assignments and assessments play an essential position since they allow us to examine student’s performance. As of the present moment, evaluation has to be performed by hand, and which requires direct, one-on-one assessment, which is affordable when working with a small class size. But manual grading becomes an overwhelming chore when the number of students grows. Furthermore, prejudice can affect human graders, leading to irregularities and mistakes in the grading process. Numerous scholars have established automated essay grading systems in an effort to overcome these issues. These algorithms take into account various factors such as syntax, semantic coherence, and the contextual use of sentiment-related terms in the text, going beyond conventional keyword based techniques. Hence, this article suggests conducting a thorough analysis of Automated Essay Scoring (AES) models. The primary goal is to elucidate the numerous methodologies, datasets, similarity metric tools, and performance metrics that can be utilized in developing AES systems. Additionally, it discusses the challenges, ethical issues, research outcomes to be considered while developing the AES system.