An Implementation of Computerized Valuation of Descriptive Answers: A Machine Learning Approach

Abstract

Evaluation is an essential part of the education and it is carried out by the system of examinations. When evaluating a big number of pupils, a significant amount of physical labor is needed. In addition to being a labor-intensive process, manual valuation varies in quality depending on the examiner's disposition. Many of the aforementioned issues would be resolved in the modern world if this could be machine controlled. Thus, utilizing computers to assess responses is one way to find a solution. However, computers still have a difficult time evaluating descriptive responses. Therefore, it is crucial to investigate and implement techniques for the automated assessment of declarative responses.
This study proposes a machine learning strategy based on classifiers for evaluating descriptive responses. We conduct an experiment in our academic institution to construct the necessary