关键词: 入侵检测 数据挖掘 规划识别 网络安全

Abstract:

This thesis presents an intelligent network-based intrusion detection system model-SmartNIDS. In this model, the network engine gathers the transferri ng network packets, and the host agent collects detailed security logs. Subsequently it describes a data mining framework for analyzing the audit record  s. Meanwhile, the model combines the plan recognition in artificial intelligence to predict the future actions of attackers and direct appropriate responses to their actions.

Key words: intrusion detection, data mining, plan recognition, network security