Pre-exam Question Paper Quality Check using Deep-Learning techniques

Abstract

This conference paper explores automating the assessment of question paper quality using Bloom's Taxonomy in education. Manual assessment is labour-intensive and error-prone, prompting innovative solutions. Techniques include natural language processing for question classification, generating questions from subjects, and gauging question complexity in community platforms. These methods aim to enhance the creation and evaluation of tests. Deep learning models like RNN, LSTM, BiRNN, and BiLSTM, combined with embedding techniques, provide accurate Bloom level categorization. Our research offers a comprehensive overview of automated methods, promising efficient and accurate assessment of educational materials, benefiting educators and students alike.