PAR: An Intelligent Tutoring System

The Practice, Assessment, and Review System is an intelligent tutoring system built to aid in the teaching and learning of course material through images and associated questions. The system works by quantifying a student’s understanding of a topic based on their responses to questions. With this information, the system can intelligently give new questions and offer help based on their previous responses.

The PAR system was originally commissioned by the Cornell Veterinary School to help students learn how to properly read horse ultrasounds. I worked with a team of 5 of my peers, led by professor Toby Dragon. I focused on designing and implementing the backend structure, which dealt with loading in question data, estimating a students knowledge of a subject, and delivering an appropriately challenging follow-up question.

Overall, this project was a great lesson in how to structure a backend system so that it can grow in complexity. Any implementational decision was discussed with future additions in mind, based on the specs of our clients. Our team was also focused on following the Scrum framework to the best of our abilities. We held daily standup meetings, kept an up-to-date scrum board, and held bi-weekly meetings with clients to show them a working prototype. It was a great way to learn proper software development practices, test-driven development, and Scrum principles all in Java. Knowing that this system will be implemented in the Cornell Veterinary School ciriculum makes it all the more meaningful.

A research poster created by coworker Kerry Buckman

Check out our Github repository to learn more!