基于数据融合的知识发现方法在网络管理中的应用

提出用于网络管理的基于数据融合的知识发现系统框架，研究数据融合技术在知识发现的数据准备和预处理阶段的应用，研究关联规则在表达网络管理知识方面的适用性并针对网络管理数据时序性的特点，引入情景规则来表示期望发掘的知识，指出网络故障管理中关联规则和情景规则的挖掘算法以及知识增量式更新的算法，并简介了原型系统的实现方法。

Data fusion supported knowledge discovery methodology for telecommunication network management

This paper presents the framework of the data fusion supported knowledge discovery system for telecommunication network. The data fusion methodology is employed in the data preparing and pretreatment stage. The association rule is used to express the knowledge in the field of network management. Furthermore, the concept of frequent episodes rule is proposed according to the time serial characteristic of network management data. The proper mining algorithms for association rule and frequent episodes rule are presented. The incremental mining algorithm is also discussed. Finally, the implementation method is simply introduced.