An advanced data leakage detection system analyzing relations between data leak activity

In order to prevent the data leakage by the internal staff, the companies protect important information of the company by inputting the behavior pattern related to the data leakage into the system in advance and by defining the employee as the staff who leaked the data, whose behavior pattern is detected when such inputted behavior pattern is detected. However, in the case of the existing system, if the data is leaked according to the pattern of the security log occurrence which is not inputted into the system, whether of the data leakage cannot be properly detected. Therefore, this study proposes a system to prevent the leakage of data in a data leakage pattern that is not input to the system by defining a set of security logs that can appear simultaneously at the time of data leakage through association analysis algorithm as a data leakage judgment scenario. As a result of experimenting the function of the system suggested, this study judged whether of data leakage with higher accuracy than the data leak detection system which does not apply association analysis algorithm, also it showed lower percentage of false positive and false Negative. This suggests that the proposed system is less likely to misjudge data leakage.