Using Text Semantic Relatedness Measures for Automatic Short Answer Grading

Automatic short answer grading is known as the task of assessing answers based on natural language automatically using computation methods and machine learning algorithms. Development of large scale smart education systems on one hand and the importance of assessment as a key factor in learning process and its confronted challenges on the other hand has significantly increased the need for an automated system with high flexibility for assessing exams based on texts. While in the process of assessing answers based on text, student s answer is compared to an ideal response and scoring is done based on their similarity semantic relatedness and similarity measures can also be employed for this aim. In this paper, several semantic relatedness and similarity measures are extensively compared in application of short answer grading. In the following, a method is proposed for improving the performance of short answer grading systems based on semantic relatedness and similarity measures which leverages students answers with the highest score as feedback. Empirical experiments have proved that using students answers as feedback can considerably improve the precision of semantic relatedness and similarity measures in automatic short answer grading.