地理信息系统中英文对照外文翻译文献.docx
- 文档编号:24139040
- 上传时间:2023-05-24
- 格式:DOCX
- 页数:23
- 大小:276.10KB
地理信息系统中英文对照外文翻译文献.docx
《地理信息系统中英文对照外文翻译文献.docx》由会员分享,可在线阅读,更多相关《地理信息系统中英文对照外文翻译文献.docx(23页珍藏版)》请在冰豆网上搜索。
地理信息系统中英文对照外文翻译文献
中英文对照外文翻译
(文档含英文原文和中文翻译)
ASurveyonSpatio-TemporalDataWarehousing
Abstract
GeographicInformationSystems(GIS)havebeenextensivelyusedinvariousapplicationdomains,rangingfromeconomical,ecologicalanddemographicanalysis,tocityandrouteplanning.Nowadays,organizationsneedsophisticatedGIS-basedDecisionSupportSystem(DSS)toanalyzetheirdatawithrespecttogeographicinformation,representednotonlyasattributedata,butalsoinmaps.Thus,vendorsareincreasinglyintegratingtheirproducts,leadingtotheconceptofSOLAP(SpatialOLAP).Also,inthelastyears,andmotivatedbytheexplosivegrowthintheuseofPDAdevices,thefieldofmovingobjectdatahasbeenreceivingattentionfromtheGIScommunity.However,notmuchhasbeendoneinprovidingmovingobjectdatabaseswithOLAPfunctionality.InthefirstpartofthispaperwesurveytheSOLAPliterature.WethenmovetoSpatio-TemporalOLAP,inparticularaddressingtheproblemoftrajectoryanalysis.Wefinallyprovideanin-depthcomparativeanalysisbetweentwoproposalsintroducedinthecontextoftheGeoPKDDEUproject:
theHermes-MDCsystem,andPiet,aproposalforSOLAPandmovingobjects,developedattheUniversityofBuenosAires,Argentina.
Keywords:
GIS,OLAP,DataWarehousing,MovingObjects,Trajectories,Aggregation
INTRODUCTION
GeographicInformationSystems(GIS)havebeenextensivelyusedinvariousapplicationdomains,rangingfromeconomical,ecologicalanddemographicanalysis,tocityandrouteplanning(Rigaux,Scholl,&Voisard,2001;Worboys,1995).SpatialinformationinaGISistypicallystoredindifferentso-calledthematiclayers(alsocalledthemes).Informationinthemescanbestoredindatastructuresaccordingtodifferentdatamodels,themostusualonesbeingtherastermodelandthevectormodel.Inathematiclayer,spatialdataisannotatedwithclassicalrelationalattributeinformation,of(ingeneral)numericorstringtype.Whilespatialdataisstoredindatastructuressuitableforthesekindsofdata,associatedattributesareusuallystoredinconventionalrelationaldatabases.SpatialdatainthedifferentthematiclayersofaGISsystemcanbemappedunivocallytoeachotherusingacommonframeofreference,likeacoordinatesystem.Theselayerscanbeoverlappedoroverlayedtoobtainanintegratedspatialview.
Ontheotherhand,OLAP(OnLineAnalyticalProcessing)(Kimball,1996;Kimball&Ross,2002)comprisesasetoftoolsandalgorithmsthatallowefficientlyqueryingmultidimensionaldatabases,containinglargeamountsofdata,usuallycalledDataWarehouses.InOLAP,dataisorganizedasasetofdimensionsandfacttables.Inthemultidimensionalmodel,datacanbeperceivedasadatacube,whereeachcellcontainsameasureorsetof(probablyaggregated)measuresofinterest.Aswediscusslater,OLAPdimensionsarefurtherorganizedinhierarchiesthatfavorthedataaggregationprocess(Cabibbo&Torlone,1997).Severaltechniquesandalgorithmshavebeendevelopedforqueryprocessing,mostoftheminvolvingsomekindofaggregateprecomputation(Harinarayan,Rajaraman,&Ullman,1996).
TheneedforOLAPinGIS
DifferentdatamodelshavebeenproposedforrepresentingobjectsinaGIS.ESRI()firstintroducedtheCoveragedatamodeltobindgeometricobjectstonon-spatialattributesthatdescribethem.Later,theyextendedthismodelwithobject-orientedsupport,inawaythatbehaviorcanbedefinedforgeographicfeatures(Zeiler,1999).TheideaoftheCoveragedatamodelisalsosupportedbytheReferenceModelproposedbytheOpenGeospatialConsortium(http:
//www.opengeospatial.org).Thus,inspiteofthemodelofchoice,thereisalwaystheunderlyingideaofbindinggeometricobjectstoobjectsorattributesstoredin(mostly)object-relationaldatabases(Stonebraker&Moore,1996).Inaddition,querytoolsincommercialGISallowuserstooverlapseveralthematiclayersinordertolocateobjectsofinterestwithinanarea,likeschoolsorfirestations.
Forthis,theyuseindexingstructuresbasedonR-trees(Gutman,1984).GISquerysupportsometimesincludesaggregationofgeographicmeasures,forexample,distancesorareas(e.g.,representingdifferentgeologicalzones).However,theseaggregationsarenottheonlyonesthatarerequired,aswediscussbelow.
Nowadays,organizationsneedsophisticatedGIS-basedDecisionSupportSystem(DSS)toanalyzetheirdatawithrespecttogeographicinformation,representednotonlyasattributedata,butalsoinmaps,probablyindifferentthematiclayers.Inthissense,OLAPandGISvendorsareincreasinglyintegratingtheirproducts(see,forinstance,MicrostrategyandMapInfointegrationinandInthissense,aggregatequeriesarecentraltoDSSs.ClassicalaggregateOLAPqueries(like“totalsalesofcarsinCalifornia”),andaggregationcombinedwithcomplexqueriesinvolvinggeometriccomponents(“totalsalesinallvillagescrossedbytheMississippiriverandwithinaradiusof100kmaroundNewOrleans”)mustbeefficientlysupported.Moreover,navigationoftheresultsusingtypicalOLAPoperationslikeroll-upordrill-downisalsorequired.TheseoperationsarenotsupportedbycommercialGISinastraightforwardway.OneofthereasonsisthattheGISdatamodelsdiscussedaboveweredevelopedwith“transactional”queriesinmind.Thus,thedatabasesstoringnonspatialattributesorobjectsaredesignedtosupportthose(nonaggregate)kindsofqueries.Decisionsupportsystemsneedadifferentdatamodel,wherenon-spatialdata,probablyconsolidatedfromdifferentsectorsinanorganization,isstoredinadatawarehouse.Here,numericaldataarestoredinfacttablesbuiltalongseveraldimensions.
Forinstance,ifweareinterestedinthesalesofcertainproductsinstoresinagivenregion,wemayconsiderthesalesamountsinafacttableoverthethreedimensionsStore,TimeandProduct.Inordertoguaranteesummarizability(Lenz&Shoshani,1997),dimensionsareorganizedintoaggregationhierarchies.Forexample,storescanaggregateovercitieswhichinturncanaggregateintoregionsandcountries.Eachoftheseaggregationlevelscanalsoholddescriptiveattributeslikecitypopulation,theareaofaregion,etc.TofulfilltherequirementsofintegratedGIS-DSS,warehousedatamustbelinkedtogeographicdata.Forinstance,apolygonrepresentingaregionmustbeassociatedtotheregionidentifierinthewarehouse.Besides,systemintegrationincommercialGISisnotaneasytask.Inthecurrentcommercialapplications,theGISandOLAPworldsareintegratedinanad-hocfashion,probablyinadifferentway(andusingdifferentdatamodels)eachtimeanimplementationisrequired,evenwhenadatawarehouseisavailablefornon-spatialdata.
AnIntroductoryExample.Wepresentnowareal-worldexampleforillustratingsomeissuesinthespatialwarehousingproblematic.WeselectedfourlayerswithgeographicandgeologicalfeaturesobtainedfromtheNationalAtlasWebsite(http:
//www.nationalatlas.gov).These
layerscontainthefollowinginformation:
states,cities,andriversinNorthAmerica,andvolcanoesinthenorthernhemisphere(publishedbytheGlobalVolcanismProgram-GVP).Figure1showsadetailofthelayerscontainingcitiesandriversinNorthAmerica,displayedusingthegraphicinterfaceofthePietimplementationwediscusslaterinthepaper.Notethedensityofthepointsrepresentingcities(particularlyintheeasternregion).Riversarerepresentedaspolylines.Figure2showsaportionoftwooverlayedlayerscontainingstates(representedaspolygons)andvolcanoesinthenorthernhemisphere.Thereisalsonon-spatialinformationstoredinaconventionaldatawarehouse.Inthisdatawarehouse,dimensiontablescontaincustomer,storesandproductinformation,andafacttablecontainsstoressalesacrosstime.Also,numericalandtextualinformationonthegeographiccomponentsexist(e.g.,population,area),storedasusualasattributes
oftheGISlayers.
Inthescenarioabove,conventionalGISandorganizationaldatacanbeintegratedfordecisionsupportanalysis.Salesinformationcouldbeanalyzedinthelightofgeographicalfeatures,convenientlydisplayedinmaps.Thisanalysiscouldbenefitfromtheintegrationofbothworldsinasingleframework.Eventhoughthisintegrationcouldbepossiblewithexistingtechnologies,ad-hocsolutionsareexpensivebecause,besidesrequiringlotsofcomplexcoding,theyarehardlyportable.Tomakethingsmoredifficult,ad-hocsolutionsrequiredataexchangebetweenGISandOLAPapplicationstobeperformed.ThisimpliesthattheoutputofaGISquerymustbeprobablyexportedasmembersindimensionsofadatacube,andmergedforfurtheranalysis.Forexample,supposethatabusinessanalystisinterestedinstudyingthesalesofnauticalgoodsinstoreslocatedincitiescrossedbyrivers.ShecouldfirstquerytheGIS,toobtainthecitiesofinterest.SheprobablyhasstoredsalesinadatacubecontainingadimensionStoreor
Geographywithcityasadimensionlevel.Shewouldneedto“manually”selectthecitiesofinterest(i.e.,theonesreturnedbytheGISquery)inthecube,tobeabletogoonwiththeanalysis(inthebestcase,anad-hoccustomizedmiddlewarecouldhelpher).Ofcourse,shemustrepeatthisforeachqueryinvolvinga(geographic)dimensioninthedatacube.
Figure1.TwooverlayedlayerscontainingcitiesandriversinNorthAmerica.
Onthecontrary,GIS/Datawarehousingintegrationcanprovideamorenaturalsolution.
Thesecondpartofthissurveyisdevotedtospatio-temporaldatawarehousingandOLAP.Movingobjectsdatabases(MOD)havebeenrec
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- 地理信息系统 中英文 对照 外文 翻译 文献
![提示](https://static.bdocx.com/images/bang_tan.gif)