基于数据流频繁模式挖掘的入侵检测模型

计算机网络入侵通常具有高频度特性,因此,识别是否正常访问,对数据流中重复元素的挖掘,给出频度指标,是一种重要的依据.提出一种基于数据流频繁模式的改进型AFP算法,该算法采用滑动窗口树技术,单遍扫描数据流及时捕获网络上的最新模式信息,并将该算法应用在入侵检测模型中正常数据和异常数据的在线挖掘.解决了有限存储和无限数据流的矛盾.实验结果表明,该模型有较高的报警率和较低的误报率.

Intrusion detection model based on frequent pattern mining over data streams

Computer network intrusion typically has the characteristics of high frequency,Therefore,the data stream mining of repeating elements and given the frequency index are the important basis for visit to identify whether it is normal.An advanced method(AFP) for mining the frequent patterns of data streams is proposed.AFP tree algorithm uses the sliding window technique.As data stream flows,the contents of the data stream are captured with a compact prefix-tree by scanning the stream only once.And the algorithm...