Research on the Impact of Attacks on Security Characteristics

摘要

Confidentiality, Integrity and Availability, essential security characteristics, are getting more attention. By analyzing the relationship between network attacks and security characteristics, we can deepen our understanding of network security. This paper proposes a method that evaluates the impact of attacks on the security characteristics for the first time by machine learning. The evaluation method is a step towards objectification. It based on intrusion detection system data set, includes data set process, extraction of critical features, evaluation value design, and other processes. This method can evaluate the impact of each attack record on the security characteristics in the data set. The evaluation result, enriching the content of the data set, can be used as a data label and an aspect of network security analysis. Through experimentation, the rationality of this evaluation method is proved.