Abstract: A data mining method based on fuzzy logic for feature selection，data analysis based on fuzzy logic reasoning for intrusion analysis，and fuzzy default logic reasoning which supports response rollback for alert response are used，which make the Intrusion Detection System（IDS） improved in the aspects of feature selection and alert response.The experiments show that with the method，intrusion attacks can be detected effectively and precisely.

摘要： 通过基于模糊逻辑的数据挖掘方法实现特征选择，使用模糊逻辑推理进行数据分析，以及支持响应回卷的模糊默认逻辑推理处理预警响应，使得入侵检测系统在特征选择和预警响应方面得到改善。实验结果显示，该检测方法能够有效检测入侵攻击，具有较低的误报率和漏报率。