Anomaly detection of access patterns in database

Data security has a critical role in the larger context of information and system security. In this paper, we propose the anomaly detection system for securing database. Our approach is based on analyzing the user´s access pattern stored in database log and detecting the anomalous access event. We consider three methods for this, user pattern analysis, machine learning analysis, and rule-based access control. Our experimental evaluation on both real and virtual database shows that our approaches work well.