基于商业智能BI的利润分析系统问题研究

江苏科技大学硕士学位论文基于商业智能（BI）的利润分析系统问题研究姓名：庹宁宏申请学位级别：硕士专业：会计学指导教师：潘燕华200803011BI12ETLOLAP34ETLOLAP52ETLOLAPIETLOLAPOLAPAbstract3ResearchonprofitanalysissystembasedonBIAbstractThepurposeofcorporateprofitanalysisistoanalyzetheraisingandfallingofprofit,thecompositionandincomeandcost.Withtheresultofprofitanalysis,wecanfindtherelationshipbetweenprofit,incomeandcost.Howtoenhancetheaccuracyandefficiencyofprofitanalysisissoimportantthateverymanagerconcernsaboutit.Today,withthehelpofinformationtechnology,lotsofMIShavetheabilityofinformationsupportingfordecisionactivities.ButMISisnotenoughtoenhancetheaccuracyandefficiency.ThemainreasonisthatprofitanalysisisbasedondifferentkindsofMIS,theseMISaredevelopedbydifferenttools,anddataareseparatedandlessofrelationship.Inordertogettheinformationthatmanagersneed,wehavetodrawdataunderthehelpofdifferentemployeesfromdifferentdepartments,sowecanimagethatthedatafromthiskindofcollectionsisnotgoodenoughorusefuletosupportdecisionmaking.Accordingtothissituation,thispapersuggestssettingupprofitanalysissystembasedonBIplatform,andthiskindofsystemcansolvesomeofthoseproblemsthatexistinprofitanalysis.Thecontentofthispaperincludesfollowingseveralaspects:1.Firstly,thispaperdescribesthereasonweusedatawarehouse,olapanddataminingtosupportprofitanalysis,anddescribestheconceptsandhistoryaboutBI,then,findthestatusaboutthisresearcharoundtheworldfromrelativematerial.2.DescribesBIanditsrelativetechnologies,includeETL,DataWarehouse,OLAP,DataMining,profitanalysissystemarebasedonthesekeytechnologies.3.Designthesystem’sstructureanddesignmethod.4.Theresearchonkeytechnologiesaboutprofitanalysissystem.First,designthedatamodelindatawarehouse;includeGranularityanddatadistrict,then,designtheETLprocess,extractionstrategy,OLAPfunctionandquerylanguage,statisticsmodelindatamining(customervalueevaluationmodel)andthemanagementaboutthemetadata,atlast,mentionedaboutsomeotherkeyproblemsinsystembuilding.5.DesignprofitanalysissystemonthebackgroundaboutFashionmanufactureThetruthtellsusthatdataanalysisbasedonBIhastheabilitytosupportconcise,comprehensiveinformation;ithelpsdecision-makingbefasterandeffective,nomatermacroormicroscopic,themanagercangetalltheinformationabouttheprofitability.Thispaperisthepreliminaryresearchtotheprofitanalysissystem,inviewofthefactthattherealisticsituationalsohadsomerestrictionfactors;itonlydiscussesthepreliminarysolution,realizationmethodandthecorrelationtechnology.Itemphasizesinthedesignsoftheconstructionandthefunction.Itwaspossibletohavesomeimperfectplace,butalsoneededthefurtherresearchandtheconsummation.Keywords:BusinessIntelligence;ETL;Datawarehouse;OLAP;DataMiningABSTRACTIIABSTRACTThepurposeofcorporateprofitanalysisistoanalyzetheraisingandfallingofprofit,thecompositionandincomeandcost.Withtheresultofprofitanalysis,wecanfindtherelationshipbetweenprofit,incomeandcost.Howtoenhancetheaccuracyandefficiencyofprofitanalysisissoimportantthateverymanagerconcernsaboutit.Today,withthehelpofinformationtechonology,lotsofMIShavetheabilityofinformationsupportingfordecisionactivities.ButMISisnotenoughtoenhancetheaccuracyandefficiency.ThemainreasonisthatprofitanalysisisbasedondifferentkindsofMIS,theseMISaredevelopedbydifferenttools,anddataareseparatedandlessofrelationship.Inordertogettheinformationthatmanagersneed,wehavetodrawdataunderthehelpofdifferentemployessfromdifferentdepartments,sowecanimagethatthedatafromthiskindofcollectiongisnotgoodenoughorusefuletosupportdecisionmaking.Concerningaboutthisreason,wesuggestdevelopprofitanalysissystembasedonBIplatform,thiskindofsystemcansolvetheproblemsthatexistinprofitanalysisnow.First,thisthesisresearchsthekeytechonologiesofBIwhichincludETL,datawarehouse,OLAP,dataminingandothers,then,basedontherequirementsofprofitanalysis,wedesignedtheconstructionofprofitanalysissystem,underthisconstruction,weresearchedthedatamodelindatawarehouse,ETLprocession,OLAPdatacubedesignandDMmodel,Atlast,thisthesismakesthedevelopmentofprofitanalysissysteminfashionmanufacureindustryanddiscusseshowtouseBItechnologiestosupportdecisionforthisindustry.Keywords:BusinessIntelligence,ProfitAnalysis,DataWarehouse,OLAP,DataMining11.1DataWarehouseOLAP,On-lineAnalyticalProcessinDataMining2OLTP,On-lineTransactionProcessingOLTP[1][2]OLAP[3]1.21.2.1CTOMagazine75%[4]3BIBusinessIntelligence[5][6]1.2.2BITPSMISDSSESBI[6]BI19921993C/SClient/Server/BrowserBI2000NBIBI1.2.31OLAPITGangadharan.G.RSwami4Sundaravalli.NBusinessIntelligenceSystems:DesignandImplementation2BIBIBIBIBIBIBI3GBIBIERPBIERPERPBIBIERPBIERPERPBIBIERPBI1.3ERPERPERPSAPERPCO-PA5HYPERIONERPOLAP1.41.4.11.1ETLOLAPETL6OLAP1.11.4.212ETLOLAP782.1OLAP[7]richdata,poorinformation1ETL[8]23[9]4[10]92.2ETLExtractionTransformationLoadOLTP1[11]ODBCFTP[12]2103[13]CPUI/O4ETLETLSASInstituteOracleWarehouseBuilderMSSQLServer2000DTSETL2.32.3.1[14]1234112.3.21[15]2[16]3[17]5~104[18]512[19]2.3.3BIETLInmonKimballInmon[20]2.1Kimball[21]Kimball[17]2.2InmonKimball2.1InmonOLAP2.2KimballOLAP13InmonKimball2.3.4;12[22]345142.4OLAPOLAPE.F.Codd1993[23]CoddOLTPSQLCoddOLAP2.4.1OLAPOLAPOLAPO