A Machine Learning-Based Automatic Feedback System to Teach Cybersecurity Principles to K-12 and College Students

Abstract

Feedback is an essential part of education to help students understand and learn from their mistakes. However, while students learn new content, there is mostly no live person to provide feedback, especially in a virtual environment. Therefore, there are many software for automated code reviews to provide feedback to programming language learners. However, there are no available auto command review tools for security tools except for each tool itself and operating system suggestions. There is also no feedback tool that constructively provides feedback according to learners’ experiences in security subjects while learners practice with commands. Therefore, we developed an automatic feedback system that uses machine learning to create customized student feedback on cybersecurity topics. The foundation of the software was completed and tested in 2 undergraduate introductory computer science courses. Survey results collected from users indicate that the automatic feedback system improved the learning experience of 46% of successful participants and that 77% of successful participants were interested in the continued development of the system. 88% of successful participants felt that the system taught basic command-line skills effectively.