Question Generation Based On Performance Analysis

Abstract

The development of an application that generates questions across various levels of Bloom's Taxonomy, such as remembrance, understanding, analysis, and creation, represents a significant advancement in the field of education and assessment. The primary goal of this application is to provide a comprehensive platform for assessing and enhancing individual performance in a specific topic. At the core of this innovative project lies the utilization of Google's state-of-the-art LLM, PaLM 2's text-bison-001, which is a powerful language model known for its natural language processing capabilities. This model is harnessed to generate questions that span the cognitive domains of Bloom's Taxonomy, ensuring a well-rounded assessment of learners' knowledge and skills based on their past performances on respective domains. There are four main modules used in the project: Providing resources, Conducting test, Performance Analysis, Question Generation. In summary, this project aspires to revolutionize education by offering personalized, comprehensive, and evolving assessments that enhance individual performance and understanding of a specific topic. Through the synergy of advanced technology and educational theory, it paves the way for a more effective and efficient learning experience.