A Feature Classification Scheme for Network Intrusion Detection

10.6633/IJNS.200707.5(1).01

Iosif-Viorel Onut；Ali A. Ghorbani

Feature classification ； feature extractor ； intrusion detection ； network security

International Journal of Network Security

5卷1期（2007 / 07 / 01）

1 - 15

英文

One of the most important phases of the IDS/IPS implementation identifies the set of features that the system is going to use. We present a feature classification schema for network intrusion detection intended to provide a better understanding regarding the features that can be extracted from network packets. Furthermore, we present the design of a feature extractor that extracts and statistically analyze features with respect to attacks. The experimental results, conducted on DARPA dataset, are intended to statistically highlight the importance of each proposed feature category, as well as to identify some of the most sensitive features to attacks.