Items Selection Strategy of Cyber Security CD-CAT Based on Collaborative Filtering

摘要

Effectively measuring personal cybersecurity literacy, accurately diagnosing personal knowledge and skills of cybersecurity, further improves personnel cybersecurity defense, the paper develops a selection strategy of adaptive cybersecurity testing based on collaborative filtering and cognitive diagnostic theory. Firstly, to better adapt to the diversified and complicated knowledge structure, the knowledge library model is designed. Then the concept of polytomous attributes is introduced to the multi-level scoring cognitive diagnosis model. And the items selection strategy based on the collaborative filtering and cognitive diagnosis is proposed for personality and commonality of the sample. Finally, experimental results show that the new strategy has a better improvement in the accuracy and safety. Comparing with the existing test strategies based on collaborative filtering or information, the developed strategy increases the effectiveness by 28.2% and 6.2% respectively, which can be effectively applied to the evaluation of personnel cyber security literacy, also providing a reference to the research.