Scholarlink

Assessing students’ educational progress crucially depends on the quality of the exam. Many educational disciplines create and review their question papers manually. This article presents an automated method for evaluating question paper quality. There are various techniques available for evaluating the quality of a question paper, and the study proposed in this article utilises Bloom’s Taxonomy for this purpose. The work utilises the BERT tokeniser and model. When constructing the dataset, university question papers from various courses are considered. The system consists of three levels: Easy, Moderate, and Hard. Based on the Bloom’s level, a score ranging from 0 to 5 is allocated, facilitating the determination of the question’s difficulty level. The findings from the experiment suggest that the proposed technology shows potential.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Difficulty Level Prediction on Evaluating the Quality of Question Papers Using Bloom’s Taxonomy

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Krish Bhikadiya,
  • Dhaval Patel,
  • Hemit Rana,
  • Nikita Bhatt

摘要

Assessing students’ educational progress crucially depends on the quality of the exam. Many educational disciplines create and review their question papers manually. This article presents an automated method for evaluating question paper quality. There are various techniques available for evaluating the quality of a question paper, and the study proposed in this article utilises Bloom’s Taxonomy for this purpose. The work utilises the BERT tokeniser and model. When constructing the dataset, university question papers from various courses are considered. The system consists of three levels: Easy, Moderate, and Hard. Based on the Bloom’s level, a score ranging from 0 to 5 is allocated, facilitating the determination of the question’s difficulty level. The findings from the experiment suggest that the proposed technology shows potential.