Abstract: For resolving cross selling which is urging needed by enterprises, this paper formulates the issue of mining multi-level temporal association rules from distributed timestamp databases, puts forward a solution for this issue. It reduces the scan times on related datasets, and also optimizes the communication mode by polling techniques for information exchange.

摘要： 为了解决目前企业急需的交叉销售问题,提出了基于时间戳数据库的分布式多层时态关联规则挖掘问题,并实现了解决该问题的核心算法DMTARM。该算法减少了数据库的扫描次数,提高了内存的使用效率和通信效率。

关键词: 数据挖掘, 关联规则, 分布式系统

CLC Number: 

• TP311