一种改进的关联规则维护算法  

An Improved Algorithm for Maintaining Discovered Association Rules

作　　者：徐文拴[1] 辛运帏[1] 

机构地区：[1]南开大学计算机科学与技术系,天津300071

出　　处：《计算机工程与应用》2006年第18期178-180,共3页Computer Engineering and Applications

摘　　要：提出了关联规则维护更新问题的背景,分析了经典的关联规则维护算法FUP2算法的局限性,针对在数据库数据集增加和删除时最小支持度同时变化的情况,提出了一种改进的关联规则维护算法EFUP,并与经典的Apriori算法进行了分析比较,表明了该算法充分利用了原有的信息,提高了算法的效率。The issue of maintaining discovered association rules is an important research topic among the various association rules mining problems.The paper explains the limitation of the classical FUP2 algorithm,proposes an improved algorithm EFUP for maintaining association rules provided that the transactions in a database is added,deleted and also the minimum support is changed,and further,compares the algorithm with the Apriori algorithm.The research result shows that the EFUP algorithm takes full advantage of former information and the rate of the algorithm is increased.

关 键 词：数据挖掘 关联规则 维护 增量更新频繁项集 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]