时态关联规则研究与应用

传统关联规则挖掘很少考虑关联规则的时间适用性.时态关联规则是指带有时态约束的关联规则,每个关联规则都有着其成立的时间区域.结合其他算法的特点,提出了基于Fisher聚类的时态关联规则挖掘算法(TApriori算法),它能够动态地发现关联规则以及关联规则有效的时间区域.实验结果证明该算法是合理有效的.最后,结合具体CRM项目的特点,讨论了在CRM引入数据挖掘的体系结构,给出了一个应用.

Research on temporal association rules and its application

Association rules discovery is a key task of data mining. But time has been little taken into account in the current research of association rules. Temporal association rule is an association rule with intervals that the rule holds. On the basis of further study of other algorithms, the paper has developed Apriori-extended TApriori algorithm, which is used to discover temporal association rules. The key idea of TApriori is using Fisher algorithm to cluster the valid time that the rules hold. Test shows its efficiency and rationality. Combined with the CRM project, the paper discusses the application of temporal association rules. And the system architecture of mining temporal association rules on CRM project has been researched.