A Comparative Analysis on Classifying Model Problems Using Revised Bloom’s Taxonomy with SVM and K-NN Techniques

Abstract

Effective education involves more than rote learning—it develops students' reasoning and creativity. The Revised Bloom’s Taxonomy (RBT) offers a structured way to assess cognitive levels, ranging from recalling facts to generating new ideas. In this study, we automate the classification of exam questions into Bloom's levels using machine learning—specifically, Support Vector Machines (SVM) and K-Nearest Neighbors (K-NN). We incorporate preprocessing steps such as grammar validation and subject-relevance checking, and leverage action verbs to map questions to cognitive levels. Our work is grounded in the findings of several studies, which indicate that SVM often surpasses K-NN in classification tasks [1–6]. We believe our approach can assist educators in designing more balanced and thought-provoking assessments.