基于两级支持度的正、负关联规则挖掘  

Approach for Mining Positive & Negative Association Rules Based on 2-level Support

作　　者：董祥军[1] 王淑静[2] 宋瀚涛[1] 

机构地区：[1]北京理工大学计算机系 [2]中国航空结算中心,北京100028

出　　处：《计算机工程》2005年第10期16-18,共3页Computer Engineering

摘　　要：当同时研究正、负关联规则时会遇到一些新问题,如非频繁项集的挖掘及如何避免产生自相矛盾的规则等。该文对这些问题进行了深入的研究,设计了能够同时挖掘频繁项集与非频繁项集以及能够同时挖掘这些项集中的正、负关联规则算法,实验表明该算法是非常有效的。A lot of new problems maybe occur when considering both positive and negative association rules simultaneously. These problems include how to mine infrequent itemsets, how to avoid generating self-contradictory rules and so on. This paper studies these problems in details. Algorithms are proposed to mine frequent and infrequent itemsets and to mine positive and negative association rules in these itemsets. Experimental results demonstrate the algorithms are efficient.

关 键 词：负关联规则 非频繁项集 相关性 支持度 置信度 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]