Practical Teaching Reform of Cybersecurity Courses Incorporating Personalized Recommendations

In recent years, with the development of cyber security, the national demand for cybersecurity-related professionals is growing, and the practice of cyber security related practices is increasing, which makes it difficult for students to find suitable practices in the massive practice resources, so the practice recommendations method for students emerges. As an emerging technique, Matrix Completion becomes one of the most powerful techniques to predict missing entries of a low-rank matrix from incomplete samples of its entries. Despite its efficiency in discovering and quantifying the interactions between students and practices, Matrix Completion in the recommender scheme suffers from the problems of low rating density and scalability. To conquer the above challenges as well as to provide quick and high-quality recommendations, we propose partition-based Matrix Completion, a novel recommender scheme that concurrently exploits locality-sensitive hash (LSH) and Matrix Completion. Specifically, taking advantage of the good properties of LSHs, our recommender scheme adopts an LSH hash table to reorder students with similar students buffered in close positions. As a result, the original student-practice rating matrix is partitioned into sub-matrices each containing a group of similar students thus having a lower rank. Matrix Completion is then used in the partitioned matrix to more effectively predict the missing ratings of a student for practice. The experimental results show that our recommendation scheme achieves better recommendation results in terms of practicing recommendation compared to other traditional recommendation methods.