MCQs Generation Using Ensemble Model for Student Performance Assessment

• Published in: Advances in Artificial Intelligence and Machine Learning Vol. 5; no. 1; pp. 3519 - 3533
• Main Authors: Madri, VijayaRaju, Meruva, Sreenivasulu
• Format: Journal Article
• Language: English
• Published: 2025
• ISSN: 2582-9793; 2582-9793
• DOI: 10.54364/AAIML.2025.51201

Multiple-Choice Questions have a vital role in the educational assessment; they are a convenient and scalable way to have students engaged in learning on multiple subjects. However, until recently, these questions had to be created almost manually by people or a team of experts who wrote questions based on the specific learning objective. As creating the question bank in this way requires a lot of effort, it is still not always feasible to use only MCQs in the assessment. The automated generation of MCQ may free some time for educators or help to evaluate the understanding of students more accurately. Recent advancements in Natural Language Processing and machine learning have made it possible to generate questions automatically based on existing educational content. This paper proposes an Time Constraint Limited MCQs Generation using Ensemble Learning Model (TCL-MCQs-ELM) that will create MCQs. The ensemble model proposed with this paper has combined the strengths of different machine algorithms.It has integrated transformer models, basic rulebased algorithms, and neural networks models, which might generate questions of low quality if used alone. This method guarantees that the generated MCQs are not just correct based on the context but also are at an acceptable rate. Other lecture note forms that would apply this ensemble model to generate MCQs are books and online-based notes. The quality assessment of the ensemble model involves carrying out an experiment pick view questions were randomly selected, and then the generated MCQ group had to select which was generated by the ensemble model. The quality, as coherence and relevancy, was compared, and the outcomes illustrated that the quality of the ensemble model is comparable to the manual MCQ generation process. Therefore, the ensemble model’s quality is considerable to be used to assess students’ knowledge of the educational platform system. However, the advantages of using this ensemble model are numerous, including scalability, reduced human input, and mediation on many educational domains, and thus apply a great range of platforms, classroom-based or e-learning.