Advanced Educational Assessments: Automated Question Classification Based on Bloom’s Cognitive Level

In the contemporary education system, the quality of question papers plays a pivotal role in evaluating students' knowledge and comprehension. To ensure the validity of assessment outcomes, it is imperative to assess the quality of these question papers, taking into account factors such as question clarity, alignment with learning objectives, structural coherence, and conformity with intended educational outcomes. This study is centered around the development of a predictive model that employs Bloom’s taxonomy—a framework for categorizing learning objectives—to gauge the difficulty level of questions. To optimize performance, we have harnessed the power of Bidirectional Long Short-Term Memory Network (BiLSTM), renowned for effectively preserving intricate dependencies within data. Through extensive experimentation on widely recognized datasets, our results showcased the superior accuracy of BiLSTM, with an overall accuracy rate of 80%, outperforming existing methods by a substantial margin of 5.44%. These findings represent a significant advancement in the realm of educational assessment, empowering educators with advanced machine learning techniques for more precise evaluation of students' cognitive capabilities.