分布式数据库全局最大频繁项集增量更新算法

随着分布式数据库记录的不断增加,需要对已挖掘出的全局最大频繁项集进行增量更新。在已经提出的快速挖掘全局最大频繁项集算法(FMMFI)的基础上,提出了分布式数据库全局最大频繁项集增量更新算法(IUGMFI)。IUGMFI算法利用了FMMFI算法已经挖掘出的分布式数据库全局频繁项目和全局最大频繁项集。挖掘增量数据库的全局频繁项目,构建增量数据库的FP-tree,挖掘出增量数据库的全局最大频繁项集,采用自上而下的剪枝策略更新全局最大频繁项集。理论分析和实验结果表明,IUGMFI算法是快速而有效的。

Incremental Updating Algorithm of Global Maximum Frequent Itemsets in Distributed Database

The global maximum frequent itemsets needed to be updated because of the increasing records in distributed database. On the basis of the fast mining algorithm for global maximum frequent itemsets, an incremental updating algorithm was proposed, namely, IUGMFI algorithm. The algorithm made use of the global frequent items and global maximum frequent itemsets in distributed database. Firstly, the global frequent items were mined in incremental distributed database. Secondly, the FP-tree was constructed in incremental distributed database. Thirdly, the global maximum frequent itemsets were mined in incremental distributed database. Finally, the global maximum frequent itemsets were updated by the strategy of top-down. Theoretical analysis and experimental results show that IUGMFI algorithm is fast and effective.