数据挖掘课件--sequence.pdfVIP

Data Mining Instructor: Associate Prof. Ying Liu Graduate University of Chinese Academy of Sciences Fictitious Economy and Data SScience Research CCenter SSequence DDattabbases SSequentitiall PPatttterns TTransactition ddattabbases, titime-seriies ddattabbases vs. sequence databases Frequent patterns vs. (frequent) sequential ppatterns Applications of sequential pattern mining CCusttomer shhoppiing sequences: • First buy computer, then CD-ROM, and then digital camera, withinwithin 33 monthsmonths. Weblog click streams DNADNA sequencessequences andand genegene structuresstructures 2010/11/2 2 PParametters DurationDuration Entire available sequence PartitionedPartitioned sequencessequences, everyevery yearyear, everyevery weekweek, everyevery twotwo weeks, etc. Event foldingg window A specified period of time, where events occurring together with no ordering, parallel episodes Interval Int = 0: no interval gap EE.gg. findfind patternspatterns inin consecutiveconsecutive weeksweeks Min_interval= int=max_interval: find patterns separated by at least min_interval at most max_interval 2010/11/2 3 WhatWhat IsIs SequentialSequential PatternPattern Mining?Mining? GivenGiven aa setset ofof sequencessequences, findfind thethe completecomplete setset ofof frequent subsequences A sequence : (ef) (ab) (df) c b A seqquence database SID sequence An element may contain a set of items. 10 a(abc)(ac)d(cf) Items within an element are unordered andd we lilistt ththem allphhabbe