【摘要】 网络流数据频繁项集挖掘是网络流量分析的重要基础。提出一种新颖的基于字典顺序前缀树LOP-Tree的频繁项集挖掘算法STFWFI,该算法采用更符合网络流特点的滑动时间衰减窗口模型,有效降低了挖掘频繁项集的时间和空间复杂度;在该树结构上提出一种新的基于统计分布的节点权值计算方法SDNW代替传统的统计方法,提高了网络流节点估值的精确度。实验结果表明,该算法在网络流频繁项集挖掘过程中获得了良好的效果。更多还原

【Abstract】 Mining network traffic frequent itemsets is an important foundation for network traffic analysis.This paper proposed a novel algorithm STFWFI(sliding time fading window frequent itemsets) based on lexicographic ordered prefix tree LOP-Tree.STFWFI used a sliding-time-fading window model which accorded with the characteristic of network traffic,and reduced the computational time complexity and space complexity efficiently.Proposed a novel node weight count measure SDNW(statistical distribution node weight)in LOP-Tree structure based on statistical distribution instead of the conventional statistical count measure,and improved the count precision of network traffic nodes.The experimental results show that STFWFI performs much better than the previous approaches in mining network traffic frequent itemsets.更多还原