Automatic Rule Generation Method for User Action Detection from Traffic Data.

User action detection is an important role in network management and security. To identify and detect user action accurately, much research has been conducted. In this paper, we propose an automatic rule generation method that improves the existing rule generation process. In our previous research, we proposed the rule-based user action detection method that shows high detection accuracy. However, there are several problems in that the rule generation process is too complicated and it takes a lot of time and effort. We conducted several comparative experiments to verify the proposed method with the previous rule generation method. Rule generation time is greatly reduced without performance degradation compared to the previous rule generation method.