Date of Award
5-2026
Document Type
Thesis
Degree Name
Bachelor of Arts (BA)
Department
Natural Sciences
First Advisor
Hamid, Fahmida
Second Advisor
Salu, Gil
Area of Concentration
Computer Science
Abstract
An increasing number of students across the globe have been utilizing Artificial Intelligence (AI) tools in their studies. These tools can be beneficial when assisting students in understanding different topics; however, hallucinations that provide incorrect explanations by these models may negatively impact the learning process. CanvasCram is proposed to create an alternative approach to generating learning materials by integrating the use of retrieval augmented generation (RAG) into the generation process for both students and teachers. RAG relies on a knowledge base, rather than solely relying on public training data that base models use to generate their output. Additionally, a Canvas API is implemented to connect the web application directly to Canvas LMS to streamline the process of adding data to the knowledge base. A Python framework, CrewAI, was chosen for supplying the RAG tool. CrewAI uses Flows and Crews to delegate tasks, such as quiz creation, to different agents who are tasked with generating the learning material. CanvasCram provides a user interface (UI) that allows users to generate quizzes and study guides with the Crews, repeatedly practice quizzes, review and interact with their study guides, view statistics on how well they are understanding the topics they have been studying, and they control what courses they wish to study within the application.
The learning materials generated by the Crews are currently targeted for users wanting to generate learning materials at a college level. We conclude by discussing the future of CanvasCram which includes deploying the application for multiple users, providing alternative forms of data for the Crews, and adding additional functionality to the UI.
Recommended Citation
Grimes, Paige, "CANVASCRAM AUTOMATING THE GENERATION OF LEARNING MATERIALS USING RETRIEVAL AUGMENTED GENERATION (RAG)" (2026). Theses & ETDs. 6955.
https://digitalcommons.ncf.edu/theses_etds/6955
Rights
The author has granted New College of Florida the nonexclusive right to archive, make accessible, and distribute for educational purposes this work in whole or in part in all forms of media, now or hereafter known. The copyright of this work remains with the author.