扩展WIT-树融合Diffset策略的频繁加权项集快速挖掘算法  

Research of fast mining algorithm for frequent weighted itemsets based on fusion of extended WIT-trees structure and Diffset strategy

作　　者：张亚梅[1] 张皓[1] 海本斋[2] 廖晓飞[3] 

机构地区：[1]河南机电高等专科学校计算机科学与技术系,河南新乡453003 [2]武汉理工大学信息工程学院,武汉430070 [3]华中科技大学计算机科学与技术学院,武汉430074

出　　处：《计算机应用研究》2015年第12期3574-3578,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(70890081);河南省科技厅基础与前沿技术研究计划资助项目(142300410334);河南省教育厅科学技术研究重点资助项目(14A520085);河南省教师教育课程改革项目(2014-JSJYYB-026)

摘　　要：针对当前算法从加权项事务数据库挖掘频繁加权项集(FWI)时效率不高的问题,提出了一种基于加权项集-Tidset树结构的FWI快速挖掘算法。首先,提出了一种加权项集-Tidset树结构;然后,使用最小加权项集阈值和向下闭合性质修剪非频繁节点;最后,利用Diffset策略允许以内存有效方式快速计算项集的加权支持度。实验结果表明,当输入数据库中FWI数较大时,提出的算法明显降低了FWI挖掘时间。相比基于先验的算法,算法平均可节省99.37%的耗时;相比基于位矩阵的加权频繁项集生成算法,提出的算法可节省99.06%的耗时,明显提升了频繁加权项集挖掘效率。To solve the poor performance of present algorithm in mining frequent weighted itemsets （FWI） from weighted items transaction databases ,this paper proposed a fast mining algorithm for FWI based on weighted itemsets-Tidset trees. Firstly, it proposed a WIT-tree structure. Then, it used minimized weighted itemsets threshold value and closed down nature to trim infrequent nodes. Finally, it used Diffset strategy to allow memory rapidly calculate the weighted support itemsets with efficient way. The experimental results show that the mining time of the proposed algorithms is significantly less than several advanced mining algorithms in mining FWI. It can save time consumption with 99.37 % comparing with the Apriori-based algorithm. It can save time consumption with 99.06% comparing with weighted frequent item sets generation algorithm based on bit matrix, which indicates that proposed algorithm has clearly improved the efficiency of frequent weighted itemsets mining.

关 键 词：频繁加权项集 数据挖掘 WIT-树 关联规则挖掘 Diffset策略 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]