Multidimensional and Hierarchical Anomaly Detection System of Web Attacks

Abstract

In recent years, the large-scale dynamic HTTP requests have raised great challenges for traditionaldetection system in web applications. In general, intrusion detection is classified into misuse detectionand anomaly detection. Misuse detection system has its own disadvantages in poor adaptabilityand high-cost of renewal and maintenance. Therefore, anomaly detection system has emergedas the improvement of misuse detection which can identify previously unknown attacks. However,learning-based anomaly detection system is prone to result in high false positives. This paper presentsa hierarchical anomaly detection system that combines multidimensional feature generating systemand classification system. The whole system is divided into three steps: firstly, construct a separatestatistical model based on large quantities of HTTP access records; secondly, adopt unsupervisedlearning algorithms to build a variety of detecting subsystems; finally, merge the results of everysubsystem by classification algorithm. We have evaluated this system by one month real web logs ofQIHU360. The results demonstrate that the proposed model has a good detection performance andtime complexity.