2024: Leveraging Large Language Models for Assisted Programming Exercise Generation

Bachelor's theses

Michael Dyer



Online educational platforms, including Coursera, Udacity, edX, and Khan Academy, offer diverse courses and rely heavily on exercises to reinforce learning. Among these, the Artemis platform from the Technical University of Munich has introduced specialized programming exercises. However, a pressing challenge is the time and effort instructors invest in creating these exercises. This bachelor thesis, supervised by Prof. Dr. Stephan Krusche and advised by Patrick Bassner, M.Sc., proposes the integration of OpenAI’s GPT-4 into Artemis to assist instructors in generating programming exercises. The solution will provide assistance in crafting problem statements and populating code repositories, enhancing both the quality and diversity of exercises. Through a methodical software engineering approach, the thesis will detail the design, implementation, and evaluation processes, with real-world feedback shaping refinements. The goal is to provide instructors with a valuable tool, ensuring that students receive high-quality and engaging content, while adapting to the evolving landscape of digital education.

Artemis is open source and available on https://github.com/ls1intum/Artemis

Head of Research Group

Prof. Dr. Stephan Krusche

Administrative office

Sophia Adelmeier

Office: 01.07.052
Phone: +49 (89) 289-18206
Fax: +49 (89) 289-18207 Boltzmannstraße 3
85748 Garching b. München