基于trie的关联规则发现算法

TrieBased algorithm of mining association rules

分析了现有的关联规则挖掘算法,总结了当前的研究概况,从数据结构的角度出发,提出了用trie做数据结构存储交易数据库的所有项集,实现快速产生频繁项集,改进关联发现的性能.该方法只需一次扫描数据库,能够支持小的支持度计数和数据库的动态修改.

Existing algorithms of mining association rules were analyzed and a new trie-based algorithm was presented from the viewpoint of data-structure.The algorithm uses trie tree to store all possible item sets and produce frequent item set fast so that the associated mining is realized.The algorithm supports small minsupp and dynamical database updating by means of single scanning of database only.