Network intrusion recognition based on cascaded conditional random fields
Yubao Shao · Journal of Jiangsu University of Science and Technology · 2008
This paper presents a network intrusion recognition method based on cascaded conditional random fields.The method can correctly and effectively recognize the hidden attack intention and plan from the large volume of low level intrusion detection system alerts in the domain of network.At first,simple intrusion recognitions are recognized by the lower model;then the result is passed to the high model to support the recognition of the complicated intrusion.It implements a self-adaptive and intelligent intrusion detection system.The experiments show that the method has the lower false alarm rate,and is better than traditional methods.