Taxonomy Based Question Generation Using Prompt Engineering for Student Assessment

The recent wide popularity of ChatGPT has resulted in immense research interest in Large Language Models in many application areas such as education, business, healthcare, and tourism. Its popularity is due to its human-like conversation and comprehension capability. In this paper, we focus on the usage of generative AI in the education system to address the challenges of generating questions to address the different levels of learning taxonomy and help educators design questions based on the different categories of students. In this work, we are trying to build an AI system that is capable of generating questions based on Bloom’s Taxonomy using prompting techniques. The system is also capable of generating personalised quizzes for students who face difficulty answering learning taxonomy-based questions. We have used the LLM API and four types of prompting techniques to generate questions in our personalised evaluation system. The paper presents the system architecture of the personalised evaluation system. The paper also presents a comparison of various prompting techniques, and the result obtained.