IM-FTS一种快速增量式频繁访问序列挖掘算法.pdf

and ComputerEngineeringApplications计算机工程与应用 IM—FTS：一种快速增量式频繁访问序列挖掘算法 彭慧丽-，张啸剑2，张亚东2 PENG Hui-lil，ZHANG Xiao-jian2,ZHANGYa-don92 1．河南广播电视大学，郑州450008 2．河南财经学院计算机科学系，郑州450002 1．HenanRadio＆TV University，Zhengzhou450008，China of Science，Henanof 450002，China 2．DepartmentComputer UniversityFinance＆Economics，Zhengzhou E-mail：ph|81@126．corn PENG incrementalfor traver- Hui-H，ZHANGXiao--jian．ZHANG Ya--dong．IM-FTS：high-speedalgorithmminingfrequent sal and Engineering sequences．Computer Applications．2009。45(3)：138-140． Abstract：Webdata intheshort FIs be whenthe databaseis growsquickly time．previousmaychanged sequence updated．Re- FISwillconsumetoomuchexecutiontimeand this extended is finding storagespace．Inpaper，an improved tostorethe results．Anefficient basedonbidirectional IM-Frsis designed previousmining algorithm constraint，calledproposed， whichutilizesthe resultsandconstrainttodiscoverthenewFTS fromtheaddedanddeletedofthe previous strategy just part database．Theand showthatIM—rrS reducesthe executiontimeand analysisexperiments algorithmefficiently average storage for f’rS． spacemining words：bidirectional traversal constraint；dwelltime；extended sequence Key lattice；frequent 摘要：由于Web数据增长迅速，先前的频繁序列随着序列库的更新而改变。若重新挖掘频繁序列会增加处理时间和数据存储 兀s，在利用先前结果和约束策略前提下，算法仅从插入和删除序列中发现新的频繁序列。分析和实验表明算法能有效缩减数据处 理时间和存储空间。 关键词：双向约束；驻留时间；扩展格；频繁访问序列 文章编号：1002—8331(2009)03—013