Automatic Classification of Questions based on Bloom's Taxonomy using Artificial Neural Network

Examination questions evaluation is critical in educational institutions because testing is one of the most prevalent techniques of evaluating students' success in a particular course. As a result, it is crucial to create a good and high-quality examination paper that caters to various cognitive levels. As a result, many instructors rely on Bloom's taxonomy proficiency level, a widely used framework for evaluating students' cognitive abilities and capabilities. According to Bloom's taxonomy, many efforts have been offered to tackle question classification automatically. The majority of existing works considered only one domain, such as computer science. This research aims to develop a classification model for classifying examination questions that fall into several domains using Bloom's taxonomy. This paper presents a method for automatically identifying questions using an Artificial Neural Network (ANN). Term Frequency - Inverse Document Frequency (TF-IDF) is used to derive the features from questions papers. We compared the ANN-based approach with Support Vector Machine based approach. According to the findings of this study, the proposed method got an accuracy of 85.2%, which is effective in classifying questions out of several domains using Bloom's taxonomy.