有关零售超市毕业设计外文翻译

有关零售超市毕业设计外文翻译毕业设计（论文）外文翻译题目对零售超市数据进行最优产品选择的数据挖掘框架：广义PROFSET模型专业网络工程外文翻译1附录英文原文ADataMiningFrameworkforOptimalProductSelectioninRetailSupermarketData:TheGeneralizedPROFSETModel1IntroductionSincealmostallmidtolargesizeretailerstodaypossesselectronicsalestransactionSystems,retailersrealizethatcompetitiveadvantagewillnolongerbeachievedbythemereuseofthesesystemsforpurposesofinventorymanagementorfacilitatingcustomercheck-out.Incontrast,competitiveadvantagewillbegainedbythoseretailerswhoareabletoextracttheknowledgehiddeninthedata,generatedbythosesystems,anduseittooptimizetheirmarketingdecisionmaking.Inthiscontext,knowledgeabouthowcustomersareusingtheretailstoreisofcriticalimportanceanddistinctivecompetencieswillbebuiltbythoseretailerswhobestsucceedinextractingactionableknowledgefromthese外文翻译2data.Associationrulemining[2]canhelpretailerstoefficientlyextractthisknowledgefromlargeretaildatabases.Weassumesomefamiliaritywiththebasicnotionsofassociationrulemining.Inrecentyears,alotofeffortintheareaofretailmarketbasketanalysishasbeeninvestedinthedevelopmentoftechniquestoincreasetheinterestingnessofassociationrules.Currently,inessencethreedifferentresearchtrackstostudytheinterestingnessofassociationrulescanbedistinguished.First,anumberofobjectivemeasuresofinterestingnesshavebeendevelopedinordertofilteroutnon-interestingassociationrulesbasedonanumberofstatisticalpropertiesoftherules,suchassupportandconfidence[2],interest[14],intensityofimplication[7],J-measure[15],andcorrelation[12].Othermeasuresarebasedonthesyntacticalpropertiesoftherules[11],ortheyareusedtodiscovertheleast-redundantsetofrules[4].Second,itwasrecognizedthatdomainknowledgemayalsoplayanimportantroleindeterminingtheinterestingnessofassociationrules.Therefore,anumberofsubjectivemeasures外文翻译3ofinterestingnesshavebeenputforward,suchasunexpectedness[13],actionability[1]andruletemplates[10].Finally,themostrecentstreamofresearchadvocatestheevaluationoftheinterestingnessofassociationsinthelightofthemicro-economicframeworkoftheretailer[9].Morespecifically,apatterninthedataisconsideredinterestingonlytotheextentinwhichitcanbeusedinthedecision-makingprocessoftheenterprisetoincreaseitsutility.ItisinthislatterstreamofresearchthattheauthorshavepreviouslydevelopedamodelforproductselectioncalledPROFSET[3],thattakesintoaccountbothquantitativeandqualitativeelementsofretaildomainknowledgeinordertodeterminethesetofproductsthatyieldsmaximumcross-sellingprofits.Thekeyideaofthemodelisthatproductsshouldnotbeselectedbasedontheirindividualprofitability,butratheronthetotalprofitabilitythattheygenerate,includingprofitsfromcross-selling.However,initspreviousform,onemajordrawbackofthemodelwasitsinabilitytodealwith外文翻译4supermarketdata(i.e.,largebaskets).Toovercomethislimitation,inthispaperwewillproposeanimportantgeneralizationoftheexistingPROFSETmodelthatwilleffectivelydealwithlargebaskets.Furthermore,wegeneralizethemodeltoincludecategorymanagementprinciplesspecifiedbytheretailerinordertomaketheoutputofthemodelevenmorerealistic.Theremainderofthepaperisorganizedasfollows.InSection2wewillfocusonthelimitationsofthepreviousPROFSETmodelforproductselection.InSection3,wewillintroducethegeneralizedPROFSETmodel.Section4willbedevotedtotheempiricalimplementationofthemodelanditsresultsonreal-worldsupermarketdata.Finally,Section5willbereservedforconclusionsandfurtherresearch.2ThePROFSETModelThekeyideaofthePROFSETmodelisthatwhenevaluatingthebusinessvalueofaproduct,oneshouldnotonlylookattheindividualprofitsgeneratedbythatproduct(thenaiveapproach),butonemustalso外文翻译5takeintoaccounttheprofitsduetocross-sellingeffectswithotherproductsintheassortment.Therefore,toevaluateproductprofitability,itisessentialtolookatfrequentsetsratherthanatindividualproductitemssincetheformerrepresentfrequentlyco-occurringproductcombinationsinthemarketbasketsofthecustomer.AswasalsostressedbyCabenaetal.[5],onedisadvantageofassociationsdiscoveryisthatthereisnoprovisionfortakingintoaccountthebusinessvalueofanassociation.ThePROFSETmodelwasafirstattempttosolvethisproblem.Indeed,intermsoftheassociationsdiscovered,thesaleofanexpensivebottleofwinewithoystersaccountsforasmuchasthesaleofacartonofmilkwithcereal.Thisexampleillustratesthat,whenevaluatingtheinterestingnessofassociations,themicro-economicframeworkoftheretailershouldbeincorporated.PROFSETwasdevelopedtomaximizecross-sellingopportunitiesbyevaluatingtheprofitmargingeneratedperfrequentsetofproducts,ratherthanperproduct.InthenextSectionwewilldiscussthelimitations外文翻译6ofthepreviousPROFSETmodel.Moredetailscanbefoundelsewhere[3].2.1LimitationsThepreviousPROFSETmodelwasspecificallydevelopedformarketbasketdatafromautomatedconveniencestores.Datasetsofthisoriginarecharacterizedbysmallmarketbaskets(size2or3)becausecustomerstypicallydonotpurchasemanyitemsduringasingleshoppingvisit.Therefore,theprofitmargingeneratedperfrequentpurchasecombination(X)couldaccuratelybeapproximatedbyaddingtheprofitmarginsofthemarketbaskets(Tj)containingthesamesetofitems,i.e.X=Tj.However,forsupermarketdata,theexisti