CAJ | 학술논문

关联规则是数据挖掘的一个重要研究内容，主要用于从大量数据集中挖掘出有价值的数据项之间的关联关系．典型案例是超市的购物篮分析，主要对顾客的购买记录数据库进行关联规则挖掘，可以发现顾客的购买行为．本文依据Apriori算法的两个基本性质，即任何大项集的子集一定是大项集，非大项集的超集一定是非大项集，对经典的Apfiori算法要多次扫面事务数据库的问题，作了一些改进，并进行仿真计算，结果表明，改进的算法确实减少了扫描次数．
관련규칙시수거알굴적일개중요연구내용，주요용우종대량수거집중알굴출유개치적수거항지간적관련관계．전형안례시초시적구물람분석，주요대고객적구매기록수거고진행관련규칙알굴，가이발현고객적구매행위．본문의거Apriori산법적량개기본성질，즉임하대항집적자집일정시대항집，비대항집적초집일정시비대항집，대경전적Apfiori산법요다차소면사무수거고적문제，작료일사개진，병진행방진계산，결과표명，개진적산법학실감소료소묘차수．
The association rule is an important rule in data mining research, mainly used to dig out the valuable relationship between data items from large data set. The typical case is a supermarket basket analysis which is mainly on the customer＇s purchase database. It can find customers＇buying behavior. Originally based on two basic natures of the Apriori algorithm, namely, any subsets of large item set must be large item set, and supersets of the set of non - major items must not be large item sets. The classical Apriori algorithm re- peatedly scans the transaction database, we make some improvements of it, and simulation results show that the improved algorithm does reduce the number of scans.