Personalized Cognitive Tutoring using Davinci-003 API for Adaptive Question Generation and Assessment

TL;DR

This paper introduces a personalized cognitive tutoring system powered by Davinci-003 API that generates and assesses adaptive questions to enhance student learning across various topics and domains.

Contribution

It presents a novel AI-driven tutoring approach that creates personalized, adaptive questions and demonstrates its feasibility through a working prototype, expanding support for diverse learners.

Findings

Prototype successfully generates personalized questions

System adapts to individual student levels

Potential for improved learning outcomes

Abstract

This paper presents a cognitive tutor powered by Davinci 003 API that generates and evaluates personalized questions for students on any topic they choose. The tutor adapts to the student's level of understanding and fosters knowledge transfer by generating questions that relate the topic to different domains. This solution has the potential to improve student learning outcomes by providing personalized and adaptive questions that challenge them at their optimal level of difficulty. The feasibility of this solution has been demonstrated through a working prototype developed using Microsoft PowerApps. Additional research could reveal how affective computing principles could be integrated into the system to analyze the emotional valence of the user and how the system could be tuned to meet the specific needs of learners across the ASD spectrum. This solution is novel and offers more comprehensive support to a diverse range of learners than existing AI tutors, such as Quizlet's Q-Chat. The paper also includes an equity statement that outlines the author's commitment to promoting educational equity and addressing potential biases in the project.