1、DecisionSupportSystemDecisionsupportsystemsareaclassofcomputerizedinformationsystemsthatsupportdecisionmakingactivities.DefinitionsTheconceptofadecisionsupportsystem(DSS)isextremelybroadanditsdefinitionsvarydependingupontheauthorspointofview(DruzdzelandFlynn1999).ADSScantakemanydifferentformsandthet

2、ermcanbeusedinmanydifferentways(Alter1980).Ontheonehand,Finlay(1994)andothersdefineaDSSbroadlyasacomputer-basedsystemthataidstheprocessofdecisionmaking.Inamorepreciseway,Turban(1995)definesitasaninteractive,flexible,andadaptablecomputer-basedinformationsystem,especiallydevelopedforsupportingthesolut

3、ionofanon-structuredmanagementproblemforimproveddecisionmaking.Itutilizesdata,providesaneasy-to-useinterface,andallowsforthedecisionmakersowninsights.Otherdefinitionsfillthegapbetweenthesetwoextremes.ForKeenandScottMorton(1978),DSScoupletheintellectualresourcesofindividualswiththecapabilitiesoftheco

4、mputertoimprovethequalityofdecisions(DSSarecomputer-basedsupportformanagementdecisionmakerswhoaredealingwithsemi-structuredproblems).ForSpragueandCarlson(1982),DSSareinteractivecomputer-basedsystemsthathelpdecisionmakersutilizedataandmodelstosolveunstructuredproblems.Ontheotherhand,Keen(1980)claimst

5、hatitisimpossibletogiveaprecisedefinitionincludingallthefacetsoftheDSS(therecanbenodefinitionofdecisionsupportsystems,onlyofdecisionsupport).Nevertheless,accordingtoPower(1997),thetermdecisionsupportsystemremainsausefulandinclusivetermformanytypesofinformationsystemsthatsupportdecisionmaking.Hehumor

6、ouslyaddsthateverytimeacomputerizedsystemisnotanon-linetransactionprocessingsystem(OLTP),someonewillbetemptedtocallitaDSS.Asyoucansee,thereisnouniversallyaccepteddefinitionofDSS.Additionally,thespecificsofitiswhatmakesitlessgeneralizedandmoredetailed.Inaddition,aDSSalsoisaspecificSoftwareapplication

7、thathelpstoanalyzedatacontainedwithacustomerdatabase.Thisapproachtocustomersisusedwhendecidingontargetmarketsaswellascustomerhabits.Asyoucanseeinthisspecificexample,itisobviousthatDSScanbeusedformorethanjustorganization.Recommendedreading:DruzdzelandFlynn(1999),Power(2000),SpragueandWatson(1993),the

8、firstchapterofPower(2002),thefirstchapterofMakaras(1999),thefirstchapterofSilver(1991),thefirsttwochaptersofSauter(1997),andHolsapleandWhinston(1996).AbriefhistoryIntheabsenceofanall-inclusivedefinition,wefocusonthehistoryofDSS(seealsoPower,2003).AccordingtoKeenandScottMorton(1978),theconceptofdecis

9、ionsupporthasevolvedfromtwomainareasofresearch:thetheoreticalstudiesoforganizationaldecisionmakingdoneattheCarnegieInstituteofTechnologyduringthelate1950sandearly1960s,andthetechnicalworkoninteractivecomputersystems,mainlycarriedoutattheMassachusettsInstituteofTechnologyinthe1960s.Itisconsideredthat

10、theconceptofDSSbecameanareaofresearchofitsowninthemiddleofthe1970s,beforegaininginintensityduringthe1980s.Inthemiddleandlate1980s,executiveinformationsystems(EIS),groupdecisionsupportsystems(GDSS),andorganizationaldecisionsupportsystems(ODSS)evolvedfromthesingleuserandmodel-orientedDSS.Beginninginab

11、out1990,datawarehousingandon-lineanalyticalprocessing(OLAP)beganbroadeningtherealmofDSS.Astheturnofthemillenniumapproached,newWeb-basedanalyticalapplicationswereintroduced.ItisclearthatDSSbelongtoanenvironmentwithmultidisciplinaryfoundations,including(butnotexclusively)databaseresearch,artificialint

12、elligence,human-computerinteraction,simulationmethods,softwareengineering,andtelecommunications.DSSalsohasaweakconnectiontotheuserinterfaceparadigmofhypertext.BoththeUniversityofVermontPROMISsystem(formedicaldecisionmaking)andtheCarnegieMellonZOG/KMSsystem(formilitaryandbusinessdecisionmaking)werede

13、cisionsupportsystemswhichalsoweremajorbreakthroughsinuserinterfaceresearch.Furthermore,althoughhypertextresearchershavegenerallybeenconcernedwithinformationoverload,certainresearchers,notablyDouglasEngelbart,havebeenfocusedonhelpingdecisionmakersinparticular.TaxonomiesAswiththedefinition,thereisnoal

14、l-inclusivetaxonomyofDSSeither.Differentauthorsproposedifferentclassifications.Attheuser-level,H?ttenschwiler(1999)differentiatespassive,active,andcooperativeDSS.ApassiveDSSisasystemthataidstheprocessofdecisionmaking,butthatcannotbringoutexplicitdecisionsuggestionsorsolutions.AnactiveDSScanbringouts

15、uchdecisionsuggestionsorsolutions.AcooperativeDSSallowsthedecisionmaker(oritsadvisor)tomodify,complete,orrefinethedecisionsuggestionsprovidedbythesystem,beforesendingthembacktothesystemforvalidation.Thesystemagainimproves,completes,andrefinesthesuggestionsofthedecisionmakerandsendsthembacktoherforva

16、lidation.Thewholeprocessthenstartsagain,untilaconsolidatedsolutionisgenerated.Attheconceptuallevel,Power(2002)differentiatescommunication-drivenDSS,data-drivenDSS,document-drivenDSS,knowledge-drivenDSS,andmodel-drivenDSS.Amodel-drivenDSSemphasizesaccesstoandmanipulationofastatistical,financial,optim

17、ization,orsimulationmodel.Model-drivenDSSusedataandparametersprovidedbyDSSuserstoaiddecisionmakersinanalyzingasituation,buttheyarenotnecessarilydataintensive.Dicodessisanexampleofanopensource,model-drivenDSSgenerator(Gachet2004).Acommunication-drivenDSSsupportsmorethanonepersonworkingonasharedtask;e

18、xamplesincludeintegratedtoolslikeMicrosoftsNetMeetingorGroove(Stanhope2002).Adata-drivenDSSordata-orientedDSSemphasizesaccesstoandmanipulationofatimeseriesofinternalcompanydataand,sometimes,externaldata.Adocument-drivenDSSmanages,retrievesandmanipulatesunstructuredinformationinavarietyofelectronicfo

19、rmats.Aknowledge-drivenDSSprovidesspecializedproblemsolvingexpertisestoredasfacts,rules,procedures,orinsimilarstructures.Atthesystemlevel,Power(1997)differentiatesenterprise-wideDSSanddesktopDSS.Enterprise-wideDSSarelinkedtolargedatawarehousesandservemanymanagersinacompany.Desktop,single-userDSSares

20、mallsystemsthatresideonanindividualmanagersPC.WhenclassifyingDSS,itcanbeviewedasverybroadorverynarrow.SinceitisdifficulttoclassifyDSSintoonlyoneclassification,thetaxonomycannotexactlybepinpointed.However,ifitisnecessary,aDSSiscertainlyclassifiedintoprecise,scientificorganizationalsoftwarethatnotonly

21、contributes,butalsoperformsdecisionmakingstepsinordertoeasethepressureforitsusers.Thefactisinafewwords,DSSisanorganizationaldecisionmakingsoftware.Otherauthors,suchasAlter,HolsappleandWhinston,DonovanandMadnick,HackathornandKeen,Golden,HevnerandPower,proposedifferenttaxonomies.Readingthefirstchapter

22、ofPower(2002)isrecommended.ArchitecturesOnceagain,differentauthorsidentifydifferentcomponentsinaDSS.SpragueandCarlson(1982)identifythreefundamentalcomponentsofDSS:(a)thedatabasemanagementsystem(DBMS),(b)themodel-basemanagementsystem(MBMS),and(c)thedialoggenerationandmanagementsystem(DGMS).Haagetal.(

23、2000)describethesethreecomponentsinmoredetail:theDataManagementComponentstoresinformation(whichcanbefurthersubdividedintothatderivedfromanorganizationstraditionaldatarepositories,fromexternalsourcessuchastheInternet,orfromthepersonalinsightsandexperiencesofindividualusers);theModelManagementComponen

24、thandlesrepresentationsofevents,facts,orsituations(usingvariouskindsofmodels,twoexamplesbeingoptimizationmodelsandgoal-seekingmodels);andtheUserInterfaceManagementComponentisofcoursethecomponentthatallowsausertointeractwiththesystem.AccordingtoPower(2002),academicsandpractitionershavediscussedbuildi

25、ngDSSintermsoffourmajorcomponents:(a)theuserinterface,(b)thedatabase,(c)themodelandanalyticaltools,and(d)theDSSarchitectureandnetwork.H?ttenschwiler(1999)identifiesfivecomponentsofDSS:(a)userswithdifferentrolesorfunctionsinthedecisionmakingprocess(decisionmaker,advisors,domainexperts,systemexperts,d

26、atacollectors),(b)aspecificanddefinabledecisioncontext,(c)atargetsystemdescribingthemajorityofthepreferences,(d)aknowledgebasemadeofexternaldatasources,knowledgedatabases,workingdatabases,datawarehousesandmeta-databases,mathematicalmodelsandmethods,procedures,inferenceandsearchengines,administrative

27、programs,andreportingsystems,and(e)aworkingenvironmentforthepreparation,analysis,anddocumentationofdecisionalternatives.Marakas(1999)proposesageneralizedarchitecturemadeoffivedistinctparts:(a)thedatamanagementsystem,(b)themodelmanagementsystem,(c)theknowledgeengine,(d)theuserinterface,and(e)theuser(

28、s).ThereareseveralwaystoclassifyDSSapplications.NoteveryDSSfitsneatlyintoonecategory,butamixoftwoormorearchitectureinone.HolsappleandWhinston(1996)classifyDSSintothefollowingsixframeworks:Text-orientedDSS,Database-orientedDSS,Spreadsheet-orientedDSS,Solver-orientedDSS,Rule-orientedDSS,andCompoundDSS

29、.AcompoundDSSisthemostpopularclassificationforaDSS.ItisahybridsystemthatincludestwoormoreofthefivebasicstructuresdescribedbyHolsappleandWhinston(1996).ThesupportgivenbyDSSerrelatedcategories(HackathornandKeen,1981):PersonalSupport,GroupSupportandOrganizationalSuppo

30、rt.Additionally,IclassifyDSSinasimilarway.ThebuildupofaDSSisclassifiedintoafewcharacteristics.1)inputs:thisisusedsotheDSScanhavefactors,numbers,andcharacteristicstoanalyze.2)userknowledgeandexpertise:Thisallowsthesystemtodecidehowmuchitisreliedon,andexactlywhatinputsmustbeanalyzedwithorwithouttheuse

31、r.3)Outputs/Feedback:Thisisusedsotheuserofthesystemcananalyzethedecisionsthatmaybemadeandthenpotentially4)makeadecision:ThisdecisionmakingismadebytheDSS,however,itisultimatelymadebytheuserinordertodecideonwhichcriteriaitshoulduse.ApplicationsAsmentionedabove,therearetheoreticalpossibilitiesofbuildingsuchsystemsinanyknowledgedomain.OneoftheexamplesisClinicaldecisionsupportsyst