Knowledge Processes and Meta Processes in Ontology

KnowledgeProcessesandMetaProcessesinOntology-basedKnowledgeManagementSteffenStaab1;2,RudiStuder1;2;3,YorkSure11InstituteAIFB,UniversityofKarlsruhe,D-76128Karlsruhe,Germanyße7,76131Karlsruhe,Germanyße,76131Karlsruhe,Germany:{staab,studer,sure}@aifb.uni-karlsruhe.detel.+49-721-6084751,fax[phone]November25th,2001KnowledgeProcessesandMetaProcessesinOntology-basedKnowledgeManagementSteffenStaab1;2,RudiStuder1;2;3,YorkSure11InstituteAIFB,UniversityofKarlsruhe,D-76128Karlsruhe,Germany2OntopriseGmbH,Haid-und-NeuStraße7,76131Karlsruhe,Germany3FZIResearchCenteronInformationTechnologies,Haid-und-NeuStraße,76131Karlsruhe,GermanyTechnologyforknowledgemanagementhassofarfocusedonthemanagementofknowledgecontainers.Wepresentanapproachthatisorientedtowardsmanagingknowledgecontentsinsteadbyidentifyingknowledgeitemsatvariouslevelsofformality.Thisisdonebyprovidingvarioustypesofmetadatathataretiedtoontologiesforconceptualinterlinkage.Knowledgeitemsareembeddedintoknowledgeprocesses,whicharesupportedbyasuiteofontology-basedtools.Inordertohandlethissortofrichknowledgeprocess,weintroduceametaprocessthatputsspecialemphasisonconstructingandmaintainingtheontologywhenintroducingknowledgemanagementsystems.Inordertoelucidateourapproach,wedescribeacasestudyaboutthebuildingofCHAR,theCorporateHistoryAnalyzeR.Keywords:Ontology;KnowledgeMangement;KnowledgeProcess;KnowledgeMetaProcess1.IntroductionInrecentyearsKnowledgeManagement(KM)hasbecomeanimportantsuccessfactorforenterprises.Increasingproductcomplexity,globalization,virtualorganizationsorcustomerorientationaredevelopmentsthataskforamorethoroughandsystematicmanagementofknowledge—withinanenterpriseandbetweenseveralcooperatingenterprises.Obviously,KMisamajorissueforhumanresourcemanagement,enterpriseorganizationandenterpriseculture—nevertheless,informationtechnology(IT)playsthecrucialenablerformanyaspectsofKM.Asaconsequence,KMisaninherentlyinterdisciplinarysubject.IT-supportedKMsolutionsarebuiltaroundsomekindoforganizationalmemorythatintegratesinformal,semi-formalandformalknowledgeinordertofacilitateitsaccess,sharingandreusebymembersoftheorganization(s)forsolvingtheirindividualorcollectivetasks[3].Insuchacontext,knowledgehastobemodeled,appropriatelystructuredandinterlinkedforsupportingitsflexibleintegrationanditspersonalizedpresentationtotheconsumer.Ontologieshaveshowntobetherightanswertothesestructuringandmodelingproblemsbyprovidingaformalconceptualizationofaparticulardomainthatissharedbyagroupofpeopleinanorganization[7].ThereexistvariousproposalsformethodologiesthatsupportthesystematicintroductionofKMsolutionsintoenterprises.OneofthemostprominentmethodologiesisCommonKADSthatputsemphasisonanearlyfeasibilitystudyaswellasonconstructingseveralmodelsthatcapturedifferentkindsofknowledgeneededforrealizingaKMsolution[10].Typically,thesemethodologiesconflatetwoprocessesthatshouldbekeptseparateinordertoachieveaclearidentificationofissues:whereasthefirstprocessaddressesaspectsofintroducinganewKMsolutionintoanenterpriseaswellasmaintainingit(theso-called”KnowledgeMetaProcess”),thesecondprocessaddressesthehandlingofthealreadysetupKMsolution(theso-called”KnowledgeProcess”)(seeFigure1).E.g.intheapproachdescribedin[9],onecanseethemixtureofaspectsfromthedifferentrolesthat,e.g.,“knowledgeidentification”and“knowledgecreation”play.Theknowledgemetaprocesswouldcertainlyhaveitsfocusonknowledgeidentificationandtheknowledgeprocesswouldratherstressknowledgecreation.Figure1.TwoorthogonalprocesseswithfeedbackloopsWestartwithdiscussingtheinteractionbetweenthetwotypesofprocessesinSection2ofthepaper.Here,wecontrastthemoredocument-orientedviewwiththemoreknowledgeitem-orientedviewandindicatetheimplicationsfortheinvolvedknowledge(meta)processes.ThenotionofknowledgeitemsisfurtherelaboratedinSection3.Specialemphasisisputonthevariousformalitylevelsoftheseknowledgeitemsandtherolemetadataplayinthiscontext.InSection4,weelaborateindetailthesecondtypeofprocess,theKnowledgeProcess.Wesetupanontology-basedapproachandidentifyfivesubprocessesthatconstitutetheKnowledgeProcess:knowledgecreation,knowledgeimport,knowledgecapture,knowledgeretrievalandaccess,andknowledgeuse.Aswewillsee,ontologiesconstitutethegluetotietogetheralltheseknowledgesubprocesses.Furthermore,ontologiesopenthewaytomoveonfromadocument-orientedviewonKMtoacontents-orientedviewofKM,whereknowledgeitemsareinterlinked,combinedandusedwithininferencingprocesses.Insection5wedefineamethodologyforintroducinganontology-basedKMsolutionintoenterprises,i.e.weaddressthefirsttypeofprocessmentionedabove,theKnowledgeMetaProcess.ThemethodologyextendsandimprovestheCommonKADSmethodology[10]byintroducing—amongothers—specificguidelinesfordevelopingandmaintainingtherespectiveontology.Specialemphasisisputonastepwiseconstructionandevaluationoftheontology.Weillustrateourconceptsandmethodsbyoutliningthedevelopmentandus