一种基于滑动窗口模型的数据流加权频繁模式挖掘方法  

A sliding window based method for weighted frequent-pattern mining over data stream

作　　者：石秀金[1] 蔡艺松 SHI Xiujin;CAI Yisong(School of Computer Science and Technology, Donghua University, Shanghai 201620, China)

机构地区：[1]东华大学计算机科学与技术学院,上海201620

出　　处：《智能计算机与应用》2018年第2期63-67,共5页Intelligent Computer and Applications

摘　　要：相对于传统的频繁模式挖掘,加权频繁模式挖掘能发现更有价值的模式信息。针对数据流中的数据只能一次扫描,本文提出了一种基于滑动窗口模型的数据流加权频繁模式挖掘方法 WFP-SW(Sliding Window based Weighted Frequent Pattern minig),算法采用WE-tree(Weighted Enumeration Tree)存储模式和事务信息,利用虚权支持度维持模式的向下闭合特性,同时获取临界频繁模式。对临界频繁模式进一步计算其加权支持度获取加权频繁模式,使得计算更新模式更加便捷。实验结果显示算法具有较高的挖掘效率并且所需的内存更少。Relative to traditional frequent-pattern mining,the weighted frequent-pattern mining can find more valuable pattern information. For the data in the data stream only scanned for one time,this paper proposes a data flow weighted frequent-pattern mining method based on sliding window. The algorithm adopts WE-tree storage mode and transaction information,and utilizes the virtual weight support to maintain the downward closing characteristic of the mode,meanwhile obtains the critical frequent mode.Furthermore,the research uses the critical frequent mode to calculate the weighted support of the mode,so as to make the computing mode and updating mode more convenient. Experimental results show that the algorithm is efficient and requires less memory.

关 键 词：事务数据流 数据流挖掘 加权频繁模式挖掘 滑动窗口模型 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]