Scaling Automated Programming Assessment Systems

The first automated assessment of student programs was reported more than 60 years ago, but this topic remains relevant and highly topical among computer science researchers teachers. In the last decade, several factors have contributed to popularity approach, such as development massive online courses, where large numbers students can hardly be assessed manually, COVID-19 pandemic with a strong presence physical relocation students, ever-increasing shortage personnel in field CS. Modern Automated Programming Assessment Systems (APASs) are nowadays implemented web applications. For applications, especially those that support immediate (on-demand) program assessments feedback, it quite challenge implement various system modules secure scalable manner. Over past six years, we developed actively deployed “Edgar”—a state-of-the-art APAS enables evaluation feedback any programming language (SQL, C, Java, etc.). article, look at application architecture focus on scalability issues. We review fundamental features dynamic analysis untrusted code execution, well complex cases static plagiarism detection, summarize lessons learned over previous research. identify challenges, show how they been addressed Edgar, then propose general architectural solutions, building blocks patterns address challenges.