Recently, a lot of work has been done in web usage mining [2]. Among
2FM-WAPMining
GivenaWAPforestssequence,tomeasurethesigni canceandfrequencyofthechangestothesupportoftheWAPs,twometrics,S-valueandF-value,areproposed.Formally,theyarede nedasfollows.De nition(S-Value)LetFwiandFw(i+1)betwoWAPforestsintheWAPhistory.LetTwkbeasubtreeofFwi,Tw(k+1)bethenewversionofTwkinFw(i+1).ThentheS-valueofTwiisde nedas
Si(Twi)=|support(Tw(k+1)) support(Twk)|
max(support(Twk),support(Tw(k+1)))
GivenathresholdαforS-value,aWAPtreeTwichangedsigni cantlyfromFwitoFw(i+1)ifSi≥α.
Heresupport(Tw(k+1))denotesthesupportvaluesofTw(k+1)inWAPforestFw(i+1).S-valueisde nedtorepresentthesigni canceofchangestothesupportvaluesofaWAPintwoconsecutiveWAPforests.GivenaWAPforestsequence,therewillbeasequenceofS-valuesforaspeci cWAPtree.ItcanbeobservedthatS-valueisbetween0and1.AlargerS-valueimpliesamoresigni cantchange.
De nition(F-Value)LetHbeaWAPhistoryand Fw1,Fw2,···,Fwn .Let S1(Twk),S2(Twk),···,Sn 1(Twk) bethesequenceofS-valuesofTwk∈Fwj,for1≤j≤n.Letαbethethreshold fortheS-value.ThentheF-valueforTwkisde nedas:F(Twk,α)=,wherefi=1,ifSi(Twk)≥α;elsefi=0.
TheF-valueforatreeisde nedtorepresentthepercentageoftimesthistreechangedsigni cantlyagainstthetotalnumberofWAPforestsinthehistory.TheF-valueisbasedonthethresholdαforS-value.ItcanalsobeobservedthatF-valueisbetween0and1.BasedontheS-valueandF-valuemetrics,frequentlymutatingwebaccesspattern(FM-WAP)isde nedasfollowing.
De nition(FM-WAP)LetHbeaWAPhistoryand Fw1,Fw2,···,Fwn .LetαandβbethethresholdsfortheS-valueandF-valuerespectively.ThenaWAPtreeTwj∈Fwm,where1≤m≤n,isaFM-WAPifF(Twj,α)≥β.
FM-WAPsrepresenttheaccesspatternsthatchangesigni cantlyandfrequently(speci edbyαandβ).FM-WAPscanbefrequentorinfre-quentaccordingtoexistingWAPminingde nitions[1].Thedi erencebetweenfrequentWAPsin[1]andtheFM-WAPsisthatfrequentWAPsarede nedbasedontheoccurrenceofWAPsintheweblogdata;whileFM-WAPsarede nedbasedonthechangepatternsofweblogdata.Thenfi
Mining web access logs... 暂无评价 8页 免费 Mining history of chan......Zhu. Mining Access Patterns Efficiently from Web Logs. In Proceedings of ...
Mining history of change... 暂无评价 3页 免费 Mining Access Pattern E ...To extract interesting or useful web access patterns, we use data mining ....
Mining web access logs using relational competitive fuzzy clustering The ...In [6], associations and sequential patterns between web transactions are ...
web access pattern tracking can be defined as web page history [3]; the mining task is a process of extracting interesting patterns in web access ...
significant changes in data Web Mining的其它定义 ...Web Mining的其他介绍 History of the Web ? 1965...to understand general access patterns and trends. ...
Discovering web access p... 暂无评价 11页 免费 Discovering Patterns in ....The next phase of web usage mining is discovering frequent user navigation ...
Abstract Efficient Mining of Temporally Annotated Sequences Sequential patterns ...pages — i.e., they follow similar sequences of web access — but with...
patterns of each class (e.g., diseased and ...mining View the debt and revenue changes by ... federal/state crime history DBs) Tools: data ...
mining View the debt and revenue changes by ... federal/state crime history DBs) Tools: data ...patterns and trends Improve the quality of ...
significant changes in data Web Mining的其它定义 ...提高挖掘效 率 Web Mining的背景 History of the ...to understand general access patterns and trends. ...
热门文档