AI-Based Question Paper Quality Assessment

DOI

https://www.doi.org/10.59256/ijsreat.20260603032

Published In

Volume 6, Issue 3 (2026)

Published Online

May-June 2026

Pages

216-221

Article Type

Original Article

Check for updates

The quality of question papers plays a pivotal role in the validity and reliability of educational assessments. Traditional peer-review and committee-based mechanisms for question paper evaluation are time-consuming, inconsistent, and prone to subjective bias. This paper proposes a comprehensive Artificial Intelligence (AI)-based framework for automated question paper quality assessment (AQPQA). The proposed system integrates Natural Language Processing (NLP), machine learning (ML), and deep learning (DL) techniques to evaluate question papers across multiple quality dimensions including cognitive level alignment (Bloom's Taxonomy), linguistic clarity, content coverage, difficulty level distribution, and subject-matter relevance. The framework employs transformer-based language models fine-tuned on a domain-specific corpus of standardized examinations. Experimental evaluations on a dataset comprising 1,200 question papers across six academic disciplines demonstrate that the proposed model achieves a classification accuracy of 91.4% for Bloom's level tagging, an F1- score of 0.88 for difficulty estimation, and a Pearson correlation of 0.93 with human expert ratings. The findings indicate that AI-driven assessment tools can significantly enhance the objectivity, efficiency, and consistency of question paper evaluation, offering scalable support to educators and examination boards worldwide.