Privacy-Centric AI Code Evaluation Assistant

In today’s educational landscape, coding and computer science have become essential skills, making efficient and secure evaluation methods more important than ever. Traditional grading methods are not only time-consuming but also present significant challenges in maintaining data privacy. With the progression of digital learning, teachers are called upon to assess a large number of students’ work, and at the same time, maintain the confidentiality of the data.

To these challenges, the team has proposed the Privacy-Centric AI Code Evaluation Assistant – an innovative solution that will revolutionize how lecturers grade coding submissions. This enhanced approach helps process code submissions rapidly and effectively and analyzes code within the PDFs and images through advanced OCR models. What makes our system unique is its strong emphasis on privacy: during the evaluation process, the AI program can identify and remove information about the student that may lead to a violation of his/her privacy rights.

This project provides a tool that is efficient, cost-effective, and more importantly respects the privacy of students and teachers to facilitate the delivery of quality feedback that enriches the learning process without compromising on privacy.