基于规则前后部约束的关联规则挖掘算法  

Mining Algorithm of Association Rules Based on Fore-part & Rear-part Constraint of the Rules

作　　者：孟月昊 王朝霞[1] 郭宇栋[1] 

机构地区：[1]后勤工程学院后勤信息与军事物流工程系,重庆401331

出　　处：《后勤工程学院学报》2017年第1期79-84,共6页Journal of Logistical Engineering University

基　　金：中国博士后科学基金项目(2014M562589);全军后勤科研计划项目(AS214R002-2)

摘　　要：目前,基于项约束的关联规则挖掘算法,未考虑用户感兴趣的规则前后部项集,常常包含了大量冗余无价值的关联规则。针对此问题,提出了一种基于规则前后部约束的关联规则挖掘算法AR_F&R。该算法根据用户需求,构造指定关联规则的前后部项集,得出针对用户需求的频繁项集和关联规则,并与具有代表性的项约束关联规则挖掘算法Recorder进行了对比实验,结果表明AR_F&R算法具有更高的挖掘效率,算法执行时间也有所降低。Current mining algorithm of association rule based on item constraints often contains a large number of redundant worthless rules for the designer's ignoring the fore-part and rear-part set of the rules in which the user is interested. In order to solve this problem, a mining algorithm called AR_F & R is proposed for association rules with constraint on fore-part and rear-part of rules. According to the needs of the users, the algorithm constructs the designated fore-part set and rear-part set of these interested rules firstly. Subsequently the frequent item sets and association rules which user needs are obtained. The contrast experiments between AR_F & R and algorithm Recorder, which is a representative item constraint association rule mining algorithm, shows that AR_F & R has a higher mining efficiency and its execution time is also reduced.

关 键 词：项约束关联规则 规则前部项集 规则后部项集 数据挖掘 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]