基于剪枝概念格模型的频繁项集表示及挖掘  

Representation and mining of frequent itemsets based on the pruned concept lattice

作　　者：胡学钢[1] 刘卫[1] 王德兴[1] 

机构地区：[1]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《合肥工业大学学报（自然科学版）》2007年第9期1095-1098,1106,共5页Journal of Hefei University of Technology：Natural Science

基　　金：国家自然科学基金资助项目(050504F);安徽省自然科学基金资助项目(050420207)

摘　　要：发现频繁项集是关联规则挖掘的关键步骤。然而,大多数频繁项集求解算法因需要产生大量候选集而降低了效率。该文在研究概念格和频繁项集关系的基础上,将剪枝概念格PCL模型引入数据库中频繁项集的表示,利用概念间的关系性质,在不丢失信息的同时能有效压缩频繁项集的规模,并提出基于PCL模型的频繁项集求解算法。该算法基于Apriori性质,在构造过程中及时、动态地剪枝,删除与频繁项集求解无关的概念,从而有效地改善了频集挖掘算法的时空性能;实验证实了算法良好的性能。Discovering frequent itemsets is a crucial step in association rule mining, but most algorithms of mining frequent itemsets scan databases many times, which decreases the efficiency. Based on the analysis of the relationship between the concept lattice and the frequent itemset, the model of the pruned concept lattice（PCL） is introduced to represent frequent itemsets in the database, and the scale of frequent itemsets is compressed efficiently. The algorithm of mining frequent itemsets based on the PCL is proposed, which prunes the infrequent concepts timely and dynamically dttring the PCL＇s construction according to the Apriori property. The efficiency of the algorithm is showh in the experiments.

关 键 词：数据挖掘 关联规则 频繁项集 概念格 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]