基于数据归约和面向属性归纳的网络流量分析系统

Network Traffic Analysis System Based on Data Reduction and Attribute-oriented Induction

为了满足网络流量分析系统对海量数据快速查询和提取的要求,主要应用数据挖掘技术中的数据归约和面向属性归纳方法,对采集到的海量数据进行了相关的前期处理,主要包括数据归约和数据泛化,使得前端系统与后台数据库间的数据交互得以高效便捷地完成。最后,利用某运营商提供的奥运专网数据对所设计的系统进行了实际测试,测试结果表明,系统运行稳定,效果良好。

In order to query and extract mass data rapidly in the network traffic analysis system, the paper executed correlative pre-processing, applying data reduction and attribute-oriented induction method of data mining technology, which includes data reduction and data generalization. It made data interaction convenient and efficient between the front-end system and background database. Finally, the system was proved to have stable performance by making use of the Olympic special network data.