分布式数据库关联规则挖掘研究

Research on Mining Association Rules in Distributed System

在分布式系统中如何挖掘关联规则是数据挖掘领域研究的一个重要课题.对关联规则分布式挖掘问题进行了深入探讨.基于以P2P网络模式构建的分布式事务数据库,对Apriori算法进行了推广.改进后的算法具有扩展性好、效率高、通信代价小和实现简单等优点.最后,还提出了一种由频繁项集高效产生强关联规则的算法.

It＇s a very important subject in the domain of data mining of how to mine association rules in distributed database system. The paper makes a thorough research on this subject. Based on distributed transactional database system constructed by P2P, the paper extends the most classical algorithm Apriori. This improved algorithm has sound extension, simple time complexity, small communication cost and simplicity. At last, the paper puts forward an efficient algorithm of getting association rules from frequent itemsets.