适合通信网的动态加权关联规则挖掘算法

A Dynamic Association Rule Mining Algorithm in Communication Networks

通信网告警相关性分析在网络故障管理中占据着重要的地位。加权关联规则挖掘是通信告警相关性分析采用的主要方法之一。然而,经典的关联规则挖掘算法在实际的网络环境中却暴露出适用性不足的缺点。本文提出了一种基于枚举树存储频繁集的关联规则挖掘算法,并结合网络动态特性与拓扑特征确定权值,最后在一个实际的网络中对该算法进行仿真,结果表明该算法具有巨大的优越性。

The alarm correlation analysis in communication Networks plays an important role in the Network fault management. The mining of weighted association rules is one of the primary methods used in communication alarm correlation analysis. However, deficiency of classic association rule mining algorithm has exposed when it is put in reality in network alarm analysis. This paper proposes to use a novel data structure based on enumeration tree to handle＇massive alarms. It also makes the weight have a strong dynamic characteristic and an ability of adapting the changing of the Networks. In the last section, the simulation experiment has indicated the huge advantage of the algorithm.