Integrating AI-Driven Collaborative Learning Tools in High Enrollment Computer Science Courses

The goal of Integrating AI-Driven Collaborative Learning Tools in High Enrollment Computer Science Courses was to provide students with personalized feedback through a new GenAI-powered learning platform.

Roy and Borela created a new GenAI-powered learning platform that includes features targeted to address instructional challenges identified by computing faculty. This work demonstrates how AI in education can be most effective when it is grounded in local pedagogy and offers a blueprint for other schools aiming to build their own content-aware GenAI tools.

Artifacts created: 

• Interview analysis 

• The GenAI-powered learning environment 

• Error identification feature 

• Tailored feedback feature  

• Inferring student knowledge feature using a Deep Knowledge Tracing model 

• Problem recommendation feature using a Dueling Deep Q-Learning model and inferred mastery profiles 

• Data analytics dashboards for faculty to monitor student engagement and identify struggling students  

• Data analytics dashboards for students to support self assessment  

• Piloting the learning environment in a large CS curse in Spring 2025, Tech Fee proposal 

970 students were impacted by the project.

This project was shared via a full paper submitted to ACM Learning @Scale conference, another paper submitted to the ICER conference, 3 posters submitted to ITiCSE and ICER, and a poster presented at Celebrating Teaching Day.