营销毕设外文文献翻译.pdfVIP

ApplicationofBayesianNetwork inImprovingCustomerCreditPrecision 1. Introduction TheconceptofCRMwas firstmentionedbyGartnerGroup in1997,itisacyclingprocess betweencorporationsand customers,which produces,collects andanalyzes customerdataandthen appliestheresultstocorporateserviceand marketactivities.CRMisabrand-newideabutnotmerely amanagement idea;itisalsoanewmanagementsystem whichdemands enterprises totransform from “product-focus”to“custom-focus”model. ThecoreofCRMisfocused oncustomers. It doessystematicresearchoncustomertoenhance customservicelevel;itsfinalgoalistoraisecustomer’ssatisfactionandloyalty,toholdformer customers, tocontend fornewcustomand opportunitiesand tobring moreprofitforcompanies. CRM’spurposeistomakecorporateacknowledgetheircustomers, todevelop itsmostpossible relationship withcustomers,torealizecustom’s biggestvalueandreach double-profitbetween corporationsand customers.Thispaperfirstintroduces methodsofcalculatingBayesian network, then establishesbiasnetwork’straditionalformationways accordingtosampledata, thentomine, analyzeand evaluatecustomersofmobiletelecommunicationinordertorealizecustomer’scredit analysis andtosupportthecorporationtomakecorrect strategicdecisions. 2. Bayesiannetwork Datamining’scalculatingmethods,beingaclassical techniqueinCRM,hasbeen respected by moreandmoreexpertsand corporations.However,calculatingmethodsoftraditionaldatamining hasitsowndefects,soweintroduceBayesiannetwork incustom creditanalysis.Bayesian network modelhascharacteristics asfollow:combingpriorinformationandsampledata,itontheonehand avoidsthesubjectiveprejudiceifonlyusingthepriorinformationand avoidstheblindsearch and wearisomecalculationiflacking ofsampledata, ontheother,itcanavoid thenoiseeffect ifonly usingposteriorinformation.Bayesiannetwork model,absorbingpriorinformation,describingthe relationship amongdatainfigureform,isavery convenient wayforforecast. 2.1 Bayesiannetwork model Bayesiannetwork isalsoc