辽宁移动经营分析系统OLAP设计与实现

辽宁移动经营分析系统OLAP设计与实现作者：孙大伟学位授予单位：北京邮电大学相似文献(10条)1.外文会议Wiwatwattana.N..Jagadish.H.V..Lakshmanan.L.V.S..Srivastava.D..CSMDLX^3:ACubeOperatorforXMLOLAPWithincreasingamountsofdatabeingexchangedandevengeneratedorstoredinXML,anaturalquestionishowtoperformOLAPonXMLdata,whichcanbestructurallyheterogeneous(e.g.,parsetrees)and/ormarked-uptextdocuments.AcoreoperatorforOLAPisthedatacube.WhiletherelationalcubecanbeextendedinastraightforwardwaytoXML,wearguesuchanextensionwouldnotaddressthespecificissuesposedbyXML.Whileinarelationalwarehouse,factsareflatrecordsanddimensionsmayhavehierarchies,inanXMLwarehouse,bothfactsanddimensionsmaybehierarchical.Second,XMLisflexible:(a)anelementmayhavemissingorrepeatedsubelements;(b)differentinstancesofthesameelementtypemayhavedifferentstructure.WeidentifythechallengesintroducedbythesefeaturesofXMLforcubedefinitionandcomputation.WeproposeadefinitionforcubeadaptedforXMLdatawarehouse,includingasuitablygeneralizedspecificationmechanism.Wedefineacubelatticeovertheaggregatessodefined.Wethenidentifypropertiesofthiscubelatticethatcanbeleveragedtoallowoptimizedcomputationofthecube.Finally,wepresenttheresultsofanextensiveperformanceevaluationexperimentgaugingthebehaviorofalternativealgorithmsforcubecomputation.2.期刊论文于波.赵征.唐世渭OLAP中的CUBE计算问题-计算机应用2003,23(1)OLAP技术是决策支持系统中的一种重要技术.CUBE计算是OLAP即席查询分析的基础.为了提高查询响应速度,需要对CUBE进行预计算.典型的CUBE计算涉及物化策略、物化路径选择与优化方法、CUBE计算方法等三个方面.文中从以上三方面出发,系统地分析和阐述了CUBE计算的概念、策略、步骤、方法以及性能等.3.外文会议MingjieXuExperimentsonRemoteSensingimagecubeanditsOLAPOLAPcananswerquestionssuchas'whatnext'and'whatif'.OLAP,whichalwaysneedsthesupportofdatawarehouse,isacomplementtodatamining.Intheearlystepsofdataminingprogress,OLAPtoolsmightbehelpfulintaskssuchasexploringthedatasets,locatingmoreimportantvariablesandunwanteddata,whichcanhelpuserstounderstandthesourcedataandquickenthedataminingprogress.InordertofinddataorinformationquicklyingreatvolumesofRemoteSensing(RS)imagedatasets,foronething,metadatabasesneedtobebuilt;foranotherthing,researchanddevelopmentarerequiredtodealwithspatialOLAPapplicationservers,whichcanmanipulatespatialdatawarehousecontainingRSimages.ThematicimagessuchasTM2,TM4,landcover,transportation,slopeandtheresultimageofERDASIMAGINEExpertClassifier,etc.,havebeeninputtedintoMSAccesstables.Ongroundsofthesetablesandtherelationalmulti-dimensionaldatamodel,dimensiontablesandthefacttablehavebeengenerated,andfurthermoreaRSimagecubestructurehasbeenconstructed.Afterthepre-computationandmaterialization,aRSimagecubehasbeencreated.ExperimentsofOLAPonthiscubehavebeencarriedout.Owningtothepre-computationandmaterialization,queriesonthecubecanbecarriedoutwithnodelay.Astheexperimentsshow,ifdatawarehouseandOLAPareadopted,notonlydifferentfactorsandtheirconcepthierarchiescanbeusedconveniently,butalsoqueriesspeedup.4.外文会议MingjieXuExperimentsonRemoteSensingimagecubeanditsOLAPOLAPcananswerquestionssuchas'whatnext'and'whatif'.OLAP,whichalwaysneedsthesupportofdatawarehouse,isacomplementtodatamining.Intheearlystepsofdataminingprogress,OLAPtoolsmightbehelpfulintaskssuchasexploringthedatasets,locatingmoreimportantvariablesandunwanteddata,whichcanhelpuserstounderstandthesourcedataandquickenthedataminingprogress.InordertofinddataorinformationquicklyingreatvolumesofRemoteSensing(RS)imagedatasets,foronething,metadatabasesneedtobebuilt;foranotherthing,researchanddevelopmentarerequiredtodealwithspatialOLAPapplicationservers,whichcanmanipulatespatialdatawarehousecontainingRSimages.ThematicimagessuchasTM2,TM4,landcover,transportation,slopeandtheresultimageofERDASIMAGINEExpertClassifier,etc.,havebeeninputtedintoMSAccesstables.Ongroundsofthesetablesandtherelationalmulti-dimensionaldatamodel,dimensiontablesandthefacttablehavebeengenerated,andfurthermoreaRSimagecubestructurehasbeenconstructed.Afterthepre-computationandmaterialization,aRSimagecubehasbeencreated.ExperimentsofOLAPonthiscubehavebeencarriedout.Owningtothepre-computationandmaterialization,queriesonthecubecanbecarriedoutwithnodelay.Astheexperimentsshow,ifdatawarehouseandOLAPareadopted,notonlydifferentfactorsandtheirconcepthierarchiescanbeusedconveniently,butalsoqueriesspeedup.5.期刊论文周国亮.冯海军.何国明.陈红.李翠平.王珊.ZhouGuoliang.FengHaijun.HeGuoming.ChenHong.LiCuiping.WangShan基于图形处理器的Cuboid算法-计算机研究与发展2009,46(z2)近年来,基于图形处理器的通用计算获得了广泛关注,并在多个领域取得了进展.内存OLAP减少了磁盘I/O,但基于单核或多核CPU的计算能力及cachemiss成为新的性能瓶颈,从而无法保证好的效率.而图形处理器由于其众多核和高带宽能够很好地适应OLAP计算特性.通过图形处理器来加速任一cuboid的计算,从而提高整个内存OLAP系统的性能.提出了基于图形处理器的分块并行算法,并对算法进行了优化及讨论了数据稀疏和数据分布倾斜等不同条件下的算法.算法通过扩展可以突破内存限制,组成磁盘、内存、显存三级流水线,适应海量数据计算;同时算法也可以作为计算整个cube的基础.通过实验比较,基于图形处理器的算法明显优于四核CPU算法.6.外文期刊AnindyaDatta.HelenThomasThecubedatamodel:aconceptualmodelandalgebraforon-lineanalyticalprocessingindatawarehousesDatawarehousingandOn-LineAnalyticalProcessing(OLAP)aretwoofthemostsignificantnewtechnologiesinthebusinessdataprocessingarena.Adatawarehousecanbedefinedasa“verylarge”repositoryofhistoricaldatapertainingtoanorganization.OLAPreferstothetechniqueofperformingcomplexanalysisovertheinformationstoredinadatawarehouse.ThecomplexityofqueriesrequiredtosupportO