基于数据挖掘的入侵检测系统设计和实现

随着网络入侵技术的不断发展, 入侵的行为表现出不确定性、复杂性、多样性等特点，
入侵检测面临许多有待解决的关键问题。本文详细介绍了基于数据挖掘的入侵检测系统的设计和具体实现，也就是用于数据预处理和分类、聚类挖掘的数据挖掘技术。在数据预处理中，我们使用基于属性抽取的方法去除干扰属性。最后，我们对系统进行了测试，通过测试结果我们发现挖掘的效率和正确率，而系统确实能够有效的检测到已知未知攻击。
关键字: 入侵检测；数据挖掘；信息安全
Abstract: As the development of the network intrusion technology, the action of intrusion
represents variable, complicated, and uncertainty characteristic. Therefore, it faces so many
problems to resolve for intrusion detection. We introduce the design and implementation of the system which adopt the Data Mine technology of data pretreatment, classify and cluster patterns mine. We use the method based on features selection to get rid of noises. Finally, we give a test to the system. Through the test, we find that our intrusion detection system is efficient to detect known and unknown attack.
Keywords: Intrusion Detection; Data Mining; Information Security