Scalable Autograder and LMS Integration

Autograder has been proven to be an effective and efficient judge in programming contests. The result can be delivered immediately. As an example, in International Olympiad in Informatics and ACM Inter-Collegiate Programming Contest, the autograder is used for judging complex programs for large numbers of contestants. Having experience with autograder in contests, we aim to use it in large programming course so that the student can have immediate result and feedback. In a large class, we have to manage hundreds of students where each student submits many small exercises regularly. Manual grading will be time consuming. However, the nature of a course is different from contest. In this research, we integrate an autograder and an existing Learning Management System (LMS) with the objective of taking the benefits from both applications. A dispatcher is designed and implemented to bridge the LMS and the autograder. LMS is used by the teachers to manage classes, scores, questions and other administrative tasks. Autograder enriches LMS functionality in order to grade programs automatically. Other than automatic grading functionality, the system must be robust, scalable and secure, that come from a large course size with an abundant amount of program submissions. The system is tested with various testing methodologies to prove the requirements are satisfied.