摘要： 现有的数据挖掘方法大致有两类：有候选项集和无候选项集，有候选项集的挖掘以Apriori算法为代表，其特点是产生大量的候选项集，重复多次扫描数据库，挖掘效率低，不适合大型数据库的挖掘。无候选项集的挖掘以FP-T方法为代表，但它不能同时挖掘多概念层的关联规则，对具有超大项ID的大型数据库，无法生成“树”结构，使用也受到限制。该文将FP-T原理引入多层关联规则的并发挖掘，通过构建一个特殊节点链的指针表，可实现超大规模数据库的并发、多层挖掘。对实现物流系统信息自动化及其它数据挖掘应用领域都具有极其重要的指导意义。

关键词: 数据挖掘, 并发挖掘, 关联规则, 物流

Abstract: Present data mining method have two kinds: one has candidate itemset generation and the other without. The former, for example Apriori algorithm, has follow some disadvantages: produce large candidate itemsets, scan database repeat, mining data inefficiently, is not suit to large database mining. The latter, for example FP-T algorithm, has follow some disadvantages: do not mine association rule of more concept level parallel, do not produce tree structure of large database. Principle of FP-T is introduced into association rule of more concept level parallel mining, by building a special note-link point table, can realize parallel mining of large database. It is of important meaning about realizing modernization of interflow of commodities and some other application respect of data mining.

Key words: Data mining, Parallel mining, Association rule, Commodities interflow

中图分类号: 

• TP311.12