外文翻譯-外文文獻(xiàn)-英文文獻(xiàn)-質(zhì)量BotC企業(yè)網(wǎng)站的綜合評(píng)估的一個(gè)索引系統(tǒng)

原文2AnIndexSystemforQualitySynthesisEvaluationofBtoCBusinessWebsiteABSTRACTItisimportantforsuccessfulelectronicbusinesstohaveahi-qualitybusinesswebsite.Soweneedanaccurateandeffectiveindexsystemtoevaluateandanalysesthequalityofthebusinesswebsite.Inthispaper,theevaluationindexsystemfollowingthe‘greybox’principleisproposedwhichconsidersbothefficiencyofbusinesswebsiteandperformanceofelectronicbusinesssystem.UsingR-Hierarchicalclusteringmethodtoextractthetypicalindexesfromsub-indexesistheoreticallyprovedtohavearationalityandeffectiveness.Finally,theevaluationmethodisbrieflydiscussed.CategoriesandSubjectDescriptorsTP393.4KeywordsBusinesswebsite;Evaluationsystem;R-Hierarchicalclustering;Systemperformance1.INTRODUCTIONBusinesswebsiteisanonlinemediabetweenbuyerandseller.Ahi-qualitywebsiteiscrucialtoacompanyforasuccessfule-business.Whatisahi-qualitybusinesswebsite?Intermsofmaintainingthewebsite,whatdowefocusonsothatthequalitymeetstheusers’needs?Apparently,usingclick-throughratetoassessthepopularitycannotobjectivelyandaccuratelyevaluatethequalityofthebusinesswebsites.Instead,weneedtorelyonscientificevaluationindexsystemandmethods.Atpresent,therearemanymethodsavailableforbusinesswebsitecomparisonorranking,suchasUsageRanking,PurchaseComparison,ExpertOpinionandSynthesisEvaluationetc.Youcanfindbothofficialauthorityandnon-governmentalorganizationthatissuetheirpowerranking.Theformeroneistomonitorandregulatethemarket,suchasCNNIC,whichorganizedthecompetitionfortheTopTenWebsitesindomestic.Thelatterone,suchasConsumerreports(www.consumerreports.org),BizRate(),ForresterResearchetc.,ismainlytoguidethewebusers’activity.Thesekindsofcomparisonorrankinghavespecialvalueingettingreputationandincreasingrecognitionofthebusinesswebsitesamongtheusers,however,e-businessenterprisecannotimprovethequalityoftheirwebsitesdirectlybasedontheresultsofthesekindsofassessments.ThemainpurposeofthispaperistodevelopanindexsystemforquantitativeevaluationoftheBtoCwebsites,whichdosenotemphasizetheincomeofthewebsitebutfocusonevaluatingofitssynthesisquality.Wehopethattheapplyingofthisindexsystemwillprovidethetechniquedevelopersandmaintainerssomereferencesfordesigning,appraisinganddiagnosingtheire-businesssystemtoimproveitsqualitylevel,andtosupportmanagerstomakedecisionsforoperationofthewebsites.2.OVERVIEWOFPREVIOUSSTUDIESComparingtothefastgrowingofe-businesswebsitesintheworld,currentlywecanrarelyfindtheparticularresearchonheevaluationindexsystemofbusinesswebsite.QEM(Thewebsitequalityevaluationmethod)proposedbyOlsinaandGodoyetc.in1999canbeconsideredasoneoftherepresentativeapproaches.Itbasedonthemainfactorstoevaluatethequalityofthewebsites,includingfunctionality(globalsearch,navigability,andcontentrelativity),usability(websitemap,addressesdirectory),efficiencyandreliability.In2000,Americanresearcher,PanlaSolaman,presentede-SERVQUALmodelbasedontheconventionalservicequalityevaluationmodelSERVQUAL.Itcontainssomefactorslikeefficiency,dealcompleteness,reliability,privacyprotection,responsiveness,recompenseandcontactetc.Inthesameyear,anotherAmericanresearcher,Hauler,introducedane-QUALmodelwhichincludesthefactorsofcontent,accessibility,navigability,designandpresentation,responsiveness,background,personalizationandcustomization,etc.In2004,F.J.MirandaGonzalezandT.M.BanegilPalaciosdevelopedanuniversalevaluationindexsystemWIS(WebAssessmentIndex)thatcanbeemployedtoassesswebsitesbydifferentorganizations.Itconsistsoffourindexesofaccessibility,navigability,speedandcontent.However,theuniversalindexsystemcannotmeasureawebsiteexactlyandabsolutelyduetotheindustryspecialty,organizationalcharacteristicsanddifferentusages.OneoftherepresentativeresearchesisMr.Itassessesthebusinesswebsitesbytestingifthedesignofthewebsitecoincideswiththeshoppingprocessofonlineconsumers.Thisstandardhasfivefactors,suchassearchandbrowse,merchandiseinformation,shoppingcart,registerandpay,serviceandsupport.Anotherindexsystemforsmallandmediumbusinesswebsitescoversthefactorsofgeneralfeatures,design,promotion,informationandtheothers.Herewelistourmajorfindingsfromthepreviousresearches:2.1UnreasonableSelectionoftheIndexSomeresearchconsidernotonlytheoriginaldesignbutalsothefactorssuchaspromotionandincomeofbusinesswebsite.Someevaluationsystemshavecorrelativeorcontradictiveindexes.Forexample,itconsidersthedownloadspeed,atthesametime,itrequiresthewebdesignersnottoexcessivelyuseflashandsoundstoslowdownthespeed.2.2UnilateralEvaluationMostoftheresearchtakestheusers’viewtoevaluatethefunctionanddesignofwebsite.Ittreatsthebusinesssystemasa‘blackbox’andignorestheimpactofsystemperformanceonthewebsitesquality.Butconsideringthefactorsofsystemperformancealoneisalsonotacompleteevaluationforimprovingservicequalityofwebsite.2.3LackofaCompleteSetofQualitySynthesisEvaluationSystemAcompletesetoftooltoevaluatethewebsitesmustincludethefollowingimportantelements:categories,factors,weights,rankingsstandardandassessmentmodel.Sofar,wehavenotseenanyliteraturediscussingcompletesetofevaluationindexsystemaimingatthequalityofBtoCwebsites.3.PRINCIPLEFORTHEQUALITYSYNTHESISEVALUATIONFirst,thethreefundamentalprinciplesweneedtofollowaretobecomprehensive,tobescientificandtobefeasible.Weshouldevaluateallthefacetsofthewebsitefromdifferentdimensionsandavoidmissingvalueofimportantfactors.Moreover,thedefinitionoftheevaluationindexshouldbeaccurate，objectiveandlogicalsoitcaneliminatetheimpactontheevaluationresultbroughtbythecorrelativeindexes.Concurrently,weneedreducethequantityofindexesoradoptthesimpleoneswhichdataiseasiertobecollected,andpreventfromcomplicatedcalculationduetotheexcessiveindexes.Themainpurposeofimprovingbusinesswebsitesistoservetheusersbetter.Theyareconcernedonlyaboutthewebsites’externalattributes,suchascontent,function,presentationandbrowsespeed,etc.So,evaluatingonlybytakingtheirviewscannotdirectlyguidetodevelop,maintainandadministratethewebsite.Justliketreatingthepatient’ssymptombutnotthediseaseitself,thetechniquedeveloperormaintainercannotradicallyimprovethequalityoftheirwebsitesbycorrectingsystemstructureandwebdesignaccordingtotheevaluationresult.Onlyafterweadoptthe‘greybox’indexsystemthatconsidersbothefficiencyofbusinesswebsiteandperformanceofe-businesssystem,wecanestablishaqualitysynthesisevaluationindexsystemtobenefitthemanagementofBtoCwebsites.4.QUALITYEVALUATIONINDEXESFORBUSINESSWEBSITESelectionofindexitemslaysdownthefoundationforconstructingevaluationindexsystem.Afterwethoroughlyanalyzetheevaluationobjectivesbasedonthecharacteristicsofbusinesswebsite,weproposeaninitialindexsystemincludes5categoriesandtotally28indexitemsshowninthefollowingTable1.Websiteself-adaptabilityreferstocapabilityofe-businesssystemintelligentlyprovidingpersonalizedserviceanddynamicoptimizingsystemperformance.SystemEfficiencyreferstotheabilitythatthesystemresponsequicklytotherequestsofnumbersofwebusers.Itcanbemeasuredthroughvaluesofsomequantitativeindexes,suchasresponsetime,throughputorutilizationrate,etc.5.OPTIMIZINGTHEEVALUATIONINDEXEItisnecessaryforourinitialevaluationsystemtooptimizeifitcanbeappliedinpractice.First,theindexesaremoreorlesscorrelativewhichwillaffectstheobjectivenessoftheevaluation.Second,therearetoomoreindexesthatwillresultinlowerefficiency.Therefore,wetrytoextractandsimplifytheindexesbyusingR-Hierarchicalclusteringmethod.Generally,Rindicatesthecoefficientofcorrelationbetweentwoitems.R-Hierarchicalclusteringmethodisusuallyappliedtoclustertheindexes.Thestepsaredescribedasfollowing.5.1CalculateCoefficientofCorrelationandClusteringItfirstlytreatseveryindexasonecluster.So,wehave28clusters.Then,coefficientofcorrelationiscalculatedbetweeneverytwoclustersbyminimum-distancemethod.Next,thetwoclusterswiththemaximalcoefficientofcorrelationareclusteredintoanewone.Thesameprocessisrepeateduntilalltheindexesareclusteredintoone.5.2AnalyzetheClusteringProcessandDetermineClustersWeanalyzethevariationofminimumcoefficientofcorrelationduringtheclusteringprocesstofindtheleappoints.Accordingtothenumberofleappointsandtheknowledgeofspecialfield,wecaneventuallydeterminehowmanyclustersweneed.ThewholeprocessisillustratedinthefollowingFigure1.FollowingtheprincipleofsimplificationandfeasibilityandconsideringthecharacteristicsofBtoCwebsite,weclusterthe28indexitemsinto10.Theprecisionrateisover0.75.5.3CalculateCorrelationIndexandExtracttheRepresentativeIndexesFirst,wecalculatethecorrelationindexthatistheaverageofRbetweenoneindexandeveryotherindexinthesamecluster.一個(gè)公式miinthisformulaisthenumberoftheindexesintheclusterthatindexXjbelongsto.Then,weselecttheindexwiththemaximalcorrelationindexinthetotal10clustersindividuallyandidentify10ofthemasthemostrepresentativeindexes.Finally,theweightsoftheindexesarederivedbytheexpertgrademethod.Thefinalindexesandtheirweightsareshowninthefollowingtable2.6.CONCLUSIONInthispaper,wehaveproposedanindexsystemforqualitysynthesisevaluationanddiagnosisoftheBtoCwebsitesfollowingthe‘GreyBox’evaluationprinciple,andscientificallydeterminedandsimplifiedtheindexitems.Usually,factoranalysisorprincipalcomponentanalysisisusedtosolvetheproblemofcommon-factorandmultipleindexes.Butthesemethodsareonlysuitableforthequantitativeindexes,andtheevaluationprocessisnottrulysimplified.Becausethenewindexisthelinearfunctionofsomeoriginalones,itstillneedstocalculatethevalueofnewindexesbycollectingallthevaluesoftheoriginalones.Inourindexsystem,mostofindexisdescriptiveone.SowehavefinalizedtheindexesbyusingtheR-Hierarchicalclusteringmethod.Itreallyhasreducedthenumberoftheevaluationindexeswithoutlosingthemajorinformationfromtheoriginalindexes.Furthermore,ithaseffectivelyavoidedtheimpactofcommon-factorsontheevaluationresult.Onlytheindexofsystemefficiencycanbemeasuredthroughquantitativesub-indexessuchasresponsetime,etc.Mostofdepictiveindexesaresubjectiveandfuzzy.Inviewofthis,weshouldusefuzzycomprehensiveanalysismethodtoevaluatetogetmoreefficiencyresult.InourfutureworkweareintendedtoproposeanevaluationmodelandconductevaluationtosomefamousdomesticBtoCwebsitestoproveifthisindexsystemisscientificandfeasible.Moreover,wewillimprovethissetofindexsystemincludingevaluationmodeltomakethewholesetofindexsystemmorefeasibility.

譯文2質(zhì)量BotC企業(yè)網(wǎng)站的綜合評(píng)估的一個(gè)索引系統(tǒng)摘要：對(duì)于成功的電子商業(yè)來(lái)說(shuō)，擁有一個(gè)高質(zhì)量的商務(wù)網(wǎng)站是很重要的。因此，我們需要一個(gè)準(zhǔn)確有效的評(píng)價(jià)指標(biāo)體系來(lái)衡量分析商務(wù)網(wǎng)站的質(zhì)量。此論文在全面考慮了商務(wù)網(wǎng)站的效率和電子商務(wù)系統(tǒng)的性能提出了基于灰盒原則建立的評(píng)價(jià)指標(biāo)體系。R型系統(tǒng)聚類的網(wǎng)站質(zhì)量評(píng)價(jià)指標(biāo)優(yōu)化方法已被證實(shí)是一種從三級(jí)指標(biāo)中提取典型指標(biāo)的合理有效方法。最后，筆者簡(jiǎn)略地討論了這種評(píng)價(jià)方式。關(guān)鍵詞：商務(wù)網(wǎng)站，評(píng)價(jià)體系，基于R型系統(tǒng)聚類的網(wǎng)站質(zhì)量評(píng)價(jià)指標(biāo)優(yōu)化，系統(tǒng)性能1.介紹商務(wù)網(wǎng)站是一個(gè)介于買賣雙方的網(wǎng)上溝通媒介。對(duì)于想要獲得成功電子商務(wù)事業(yè)的公司來(lái)說(shuō)，一個(gè)高質(zhì)量的網(wǎng)站是十分關(guān)鍵的。怎么才是一個(gè)高質(zhì)量的商務(wù)網(wǎng)站呢？就維持運(yùn)行網(wǎng)站而言，為滿足用戶的需要，我們最需要關(guān)注的是什么？顯然，僅通過(guò)點(diǎn)擊率評(píng)價(jià)一個(gè)網(wǎng)站的受歡迎度并不是一個(gè)客觀精確的方法來(lái)衡量商務(wù)網(wǎng)站的質(zhì)量。反而，我們需要的是科學(xué)的評(píng)價(jià)指標(biāo)體系和方法?，F(xiàn)下，有很多可利用的方式來(lái)比較與排名商務(wù)網(wǎng)站，例如使用排名，購(gòu)買比較，專家意見(jiàn)，綜合評(píng)估等等。既有官方機(jī)構(gòu)又有非官方組織發(fā)布了他們的實(shí)力排行。前者主要是監(jiān)督調(diào)節(jié)市場(chǎng)，比如中國(guó)互聯(lián)網(wǎng)絡(luò)信息中心（其在國(guó)內(nèi)組織競(jìng)賽評(píng)選十大影響力網(wǎng)站）。而后者的功能是引導(dǎo)網(wǎng)站用戶的活動(dòng)。例如Consumerreports(www.consumerreports.org),BizRate(),ForresterResearch等等。這種類型的比較和排名能夠幫助商務(wù)網(wǎng)站在用戶中樹(shù)立名望和增加認(rèn)知度，但是電子商務(wù)公司并不能直接依靠這些評(píng)估結(jié)果來(lái)提高他們的網(wǎng)站質(zhì)量。本文的主要目的就是為電子零售網(wǎng)站的定量評(píng)估建立一個(gè)指標(biāo)體系，這個(gè)體系并不是強(qiáng)調(diào)網(wǎng)站的收入高低，而是關(guān)注網(wǎng)站的綜合質(zhì)量。我們希望通應(yīng)用這個(gè)指標(biāo)體系，可以促進(jìn)技術(shù)發(fā)展以及提供一些參考用于網(wǎng)站的設(shè)計(jì)，評(píng)估和診斷，以此達(dá)到提高網(wǎng)站質(zhì)量水平，幫助管理者抉擇網(wǎng)站的運(yùn)行。2.早期研究概要相較于當(dāng)今世界上的電子商務(wù)的高速發(fā)展，現(xiàn)今我們幾乎很難找到商務(wù)網(wǎng)站的評(píng)價(jià)指標(biāo)體系的特定研究。由Olsina和Godoy等人于1999年提出的網(wǎng)站質(zhì)量評(píng)估方法可被認(rèn)為一種較為典型的評(píng)估方式。這個(gè)評(píng)估方式主要從函數(shù)性（全球研究，可靠性和內(nèi)容相對(duì)論），可用性（網(wǎng)站視圖，地址目錄），效能和可靠性等作為主要因素來(lái)評(píng)價(jià)網(wǎng)站的質(zhì)量。2000年，美國(guó)學(xué)者PanlaSolaman，基于傳統(tǒng)的服務(wù)質(zhì)量評(píng)估模型“服務(wù)質(zhì)量”提出的“電子服務(wù)質(zhì)量”模型。其包括了一些因素，例如效能，協(xié)議完整性，可靠性，隱私保障，回應(yīng)性，賠償和接觸等。同一年，另一位美國(guó)研究者Hauler提出了另一種電子服務(wù)質(zhì)量模型，其中包括了內(nèi)容，可達(dá)性，可航性