基于加权构造链表的频繁加权项集挖掘算法  

Weighted B-list-based algorithm for mining frequent weighted itemsets

作　　者：文凯[1,2,3] 许萌萌 耿小海 WEN Kai;XU Meng-meng;GENG Xiao-hai(School of Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;Research Center of New Telecommunication Technology Applications,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;Chongqing Information Technology Designing Limited Company,Chongqing 401121,China)

机构地区：[1]重庆邮电大学通信与信息工程学院,重庆400065 [2]重庆邮电大学通信新技术应用研究中心,重庆400065 [3]重庆信科设计有限公司,重庆401121

出　　处：《计算机工程与设计》2020年第7期1920-1925,共6页Computer Engineering and Design

摘　　要：针对当前加权频繁项集挖掘算法建树复杂、挖掘效率较低的问题,提出基于加权构造链表(WB-list)的加权频繁项集挖掘BFWI算法。构造高度压缩信息的加权构造树(WB-tree),由B-list扩展结构WB-list得到节点信息,以集合枚举树作为搜索空间,结合包含索引减少项集连接次数并利用超集等价性质加快加权频繁项集的产生,提高算法的效率。实验结果表明,BFWI算法在时间和空间效率性能上优于IWS和WIT-FWIs-Diff算法,无论是处理稀疏数据还是稠密数据均得到良好效果。Aiming at the complexity of tree construction and the algorithm inefficiency of mining frequent weighted itemsets from weighted database,BFWI algorithm was proposed and it was a mining algorithm of frequent weighted itemsets based on weighted building list(WB-list).A weighted building tree(WB-tree)of highly compressed information was constructed and the node information was obtained by WB-list,an extension of B-list.Besides,the set enumeration tree was used as the search space and the number of itemsets connection was reduced by combining with the subsume index.An algorithm was built based on the equivalence property of superset for efficiently mining frequent weighted itemsets.Experimental results show that BFWI algorithm is superior to IWS and WIT-FWIs-Diff algorithms in time and space efficiency,and it has good performance in dealing with both sparse and dense data.

关 键 词：加权构造链表 频繁加权项集挖掘 超集等价 包含索引 数据挖掘 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]