改进的模糊关联规则及其挖掘算法

Improved fuzzy association rule and its mining algorithm

针对现有网络事件的关联规则未考虑到事件发生频度,导致不能完整准确反映事件关联关系的问题,采用模糊理论将事件发生频度引入到关联规则中,定义网络事件之间改进的模糊关联规则;在传统遗传算法的基础上,引入并改进兴趣度和相似度的概念,采用小生境技术调整适应度函数,提出一种基于改进模糊遗传算法的网络关联规则挖掘方法。实验结果表明,改进的模糊关联规则显著拓宽了关联规则的内涵及其挖掘范围,降低了关联规则冗余度;所提挖掘方法具有一定效率优势。

Aiming at the problem that traditional association rules fail to reflect the association relations of events without considering occurrence frequency,fuzzy theory was adopted to introduce the occurrence frequency into association rules,and the fuzzy association rules of network events were defined.Meanwhile,interesting rate and similarity rate were introduced and improved,and little world technology was adopted to adjust applicability,then an improved genetic algorithm（GA）based network association rules mining algorithm was proposed.Experimental results show new association rules can extend their meaning and mining range,reduce the redundancy and improve the efficiency.