外文翻譯--關于萬維網新時代的學報 英文版.pdf

WebSemantics:Science,ServicesandAgentsontheWorldWideWeb1(2003)15EditorialAnewjournalforaneweraoftheWorldWideWebWearedelightedtowelcomeyoutothefirstissueoftheJournalofWebSemantics.Withyourhelpweaimtomakethisjournalthepremierpublicationforaneweraofcomputing:oneinwhichmachine-readablesemanticsenableanintelligentlycapableWeb.The“SemanticWeb”isthemostwellknownversionofthisnewvision,and,despiteitsrelativeyouth,hasal-readypromotedaflurryofaction.Fromexcitingnewresearchtothedeploymentofindustrialstandards；fromacademicexperimentalprototypestocommer-cialendeavours:weareatthecentreofamaelstromofactivity.ThelanguagesneededtodefinetheSemanticWeb；thearchitecturalcomponentsandtoolsneededtobuildandmaintainit；thecontentnecessarytouseit；andtheapplicationsthatwillexploititalltheseactivitiesarehappeningatonceandyetareinterdepen-dent.ThismakestheSemanticWebanexcitingplace.First,afewquestions:Whatare“WebSemantics”,whattechnologiesdoweneedtodeliversemanticstotheWebandhowmighttheybeusedbyWeb-basedapplications?Giventhewiderangeandrelativematurityofac-tivitiesinthecommunity,howwillthisjournalprovidethebreadthanddepthneeded,andhowwillititselfbecomepartoftheSemanticWeb?HowdoesthisfirstissuereflecttheambitionsoftheSemanticWebandsetthetonefortheJournalofWebSemantics?Now,afewanswers.1.Science,servicesandagentsontheWorldWideWebTheurgetofind,collect,storeandshareinformationhasalwaysbeenwithmankind.TheWebhasmadethiseasierthanever.Ithasrevolutionizedthewayweseekinformation.Ithasbroughtdemocracytopubli-cation.Ithasspeededupthedisseminationoffacts,aswellasfictions,toaglobalcommunity.Itoffersaubiquitousinterfacetodatabasesanddocumentman-agementsystemsandauniversalconnectivefabricforintranetsaswellastheInternet.Thegoodnewsisthatifyouneedapieceofin-formationitissuretobeavailabletoyousomewhere.Theproblemishowtofinditandhowtointegratedifferentpiecesinameaningfulway.Documentman-agementsystemsandsearchenginesdonotprovideanswerstheyoffermoreorlessrelevantdocumentstobeinterpretedbythehumanreader.Aquerytoadatabaseonlyprovidesexactanswersandcannotsug-gestresultsbeyonditscurrentcontent.Tosearchforandlinkinformation,apersonorsomespecificapplicationmustinterpretthecontentoftheseinformationresources.Tomakethecontentsofdocumentsandthelinksbetweenthemgener-allymachine-interpretable,tomakethecontentsofdatabasesinterpretableonaconceptuallevel,wemustassociatewithwebresourcesmetadatathatconveystheirsemanticshencetheSemanticWeb1.TheSemanticWebdoesnotreplacetheWeb；itoffersanintegratingdescriptivefabricalongsidethewebforsearchengines,informationbrokersandultimatelyintelligentagents.NoonetechnologyholdsthemonopolyofWebSemantics.Forexample:Underpinningmetadatawithpreciseandsharedse-manticsrequiresontologiestoprovideaconsensual,sharedconceptualizationofadomainbasedonaconsensusbuildingprocess(see2andDhisissue).1570-8268/$seefrontmatter2003ElsevierB.V.Allrightsreserved.doi:10.1016/j.websem.2003.09.0022Editorial/WebSemantics:Science,ServicesandAgentsontheWorldWideWeb1(2003)15Ontologiesrelyonformalknowledgerepresentationlanguagesthatintegrateaspectsofformallanguageswiththerequirementsoftheweb.TheWebOn-tologyWorkingGroupofW3C,recentlyproposedOWLastheontologylanguagefortheSemanticWeb(seeHhisissue).WebServicesbringacomputationalelementforaccessingandexecutingsoftwarecomponentsandapplications.DevelopmentssuchasDAML-S3andTheWebServiceModelingFramework4aimatintegratingSemanticWebmethodswithWebServices,toenableautomaticservicediscovery,configurationandexecution(seeShisissue).Agentsbenefitfromthedeclarativeframework；agent-basedsystemswillevolveintoeffectivesys-temsoncemoremachine-interpretablecontentandintelligentservicesareavailableontheweb5.Databaseviewmanagement,schematransforma-tion,schemaintegration,andqueryprocessingofferaplethoraofexperienceinscaleablesemantictechnologies(seeMhisissue).Com-binedwiththestrengthsintransactionsaspectsandscalability,thedatabaseareawillbeanimportantcontributortothefurtherdevelopmentofSemanticWebapplications6.Thewebisnottheonlydistributedcomputinginfrastructurethatcanbenefitfromsemantics.Wearebeginningtoseetheintegrationofsemanticas-pectsintoPeer-to-PeerSystemsandtheGrid.Peerselectionormessageroutingcanbeoptimizedbyhavingmoresemanticinformationavailableabouttheservicesapeeroffersortheinformationapeerisstor-ing(seeAhisissue).ThemergingofGridcapabilitieswithWebServices(theOpenGridServiceArchitecture)anddevelopmentsinSemanticGridsenablethedynamicformulationof“VirtualOrganizations”ofGridresourcesandtheintegra-tionofdatafromdifferentsourcesinasemanticallyconsistentway7,8.Manyapplicationareasandindustrysectorsal-readybenefit,orwillcometodoso,fromthenewsemanticinfrastructurethatevolves.e-Commerce,orEnterpriseApplicationIntegration,gainanewlevelofflexibilityforrunningbusiness-to-businessapplicationsornetworkedenterprises.e-Science,ex-ploitingsemanticgridtechnologies,willallownewwaysofcooperationamongscientistsandthusenableanewlevelofsynergybetweenresearchersworkingindifferentinstitutionsandlocations9.Newgen-erationknowledgemanagementsolutionsinwhichknowledgemanagementisaneffortlesspartofdaytodayactivities,andwhereappropriateknowledgeisautomaticallydeliveredtotherightpeopleattherighttimeattherightgranularityviaarangeofuserdevices,isanotherpromisingapplicationarea.Thewebandweb-basedapplicationswillreachanewleveloffunctionalityonlyifwebcontentsandWebServicesarecharacterizedinawaythatdeliversasmuchsemanticsasisneededtomeettheapplicationneeds.2.StructureandcontentsofthejournalNoonedisciplineholdsthemonopolyofWebSe-mantics.Distributedcomputing,dataandknowledgemanagement,artificialintelligence,digitallibraries,languagedesignandimplementation,architecture,naturallanguageprocessingallplaytheirrole.Itistheconfluenceofthesetechnologiesthatisthekey,andsoitisthekeyfortheJournalofWebSe-mantics.Weplantobringtogetherthebestresearchfromalldisciplinesaimedatcapturingandexploitingsemanticsindistributedinformationmanagement.Theresearchisdrivenbyapplicationsandexposedthroughdemonstrators；experimentalprototypesarerunningalongsidecommercialdevelopments；funda-mentalresearchisworkinginconsortwithstandardsactivitiesinW3Candotherbodies.Thejournalwillreflectallthesestreamsofactivityandtheirinteractiontogiveawindowonthewholeofthestateoftheart.Weaimtocoveradynamicandvibrantnewarea,andtargetanaudiencethatneedstoknowthelatestinnovationswhiletheyarefresh.Weplantoachievethiswithoutsacrificingqualitybyprovidingamixoftraditionalresearchpapersaswellashigh-qualitylettersandshortarticlesthatpresentimportantresultswiththeshortestpossiblepublicationdelays.Thus,therangeofpaperswillreflectthediversityofactivity,andmaturity:Papers,ShortorLong:Thesearetraditionalre-searchpapersdescribingnovelresearch,large-scaleexperiments,orexcitingsystemsofrelevancetothejournalsreadership.Wehavenospecificpaperlimit；Editorial/WebSemantics:Science,ServicesandAgentsontheWorldWideWeb1(2003)153weparticularlyencourageshortpapersonlessmaturebutexcitinginnovations.Letters:Lettersareonetotwopagenotifications(ofthekindfoundinScientificpublicationslikeScience)focusingonaspecificresultorimportantinnovation,theoreticalorpractical.Letterswillbeanimportantwayforexcitingresultstogetintorefereedpublicationintheshortestpossibletime.Demonstrations:Demonstrationpapersareshortpapersdescribingafreelyavailabledemonstration,accompaniedbyapointertoasitewherethedemon-strationruns,orfromwhichitcanbedownloaded.Reviewerswillcheckthequalityofthedemonstra-tion,andthatthepaperpresentsenoughinformationtounderstandwhatitdoesandthatitactuallyworks.Ourwebsitegivesdetailsoflicensingarrangements.Ontologies:Ontologypapersareexpectedtodescribethedevelopmentofapubliclyavailableon-tology,whatisinit,andwhyitisimportant.Theontologyshouldbepublishedinastandardlanguageorthatthedetailsofitsrepresentationshouldbepub-lishedandavailabletoourreaders.Wewillreviewthequalityofthewrite-up,andthemodelingqualityoftheontologyitself.Youcanhelpusmakethisjournalrelevanttoyourneeds.Ifyouhavesomethingyouthinkbelongsinthejournal,butitdoesnoteasilyfitintooneofthesecategories,contactoneoftheEditors-in-Chief.3.PracticingwhatwepreachWeexpectthisjournaltohaveastrongwebpresencewherewepracticewhatwepreach.Thusweplantousewebtechnologiesinanumberofdifferentways:Rapidpublicationmixinge-journalandprintjour-nalmodels.Finalsubmissionswillbepostedonourwebsiteimmediatelyafteracceptance.Aftertypeset-tingtheywillappearintraditionalprintform,andonElsevierswebsiteandintheirelectronicarchivesamajorresourceavailableinthousandsoflibrariesaroundtheworld.Metadatamark-upusingDublinCorewillbecre-atedforyou,andyouwillhavetheopportunitytoaddanyothermetadatayouwishusingtoolsdevel-opedbythecommunity.Ourpaperrepositorywillbedisplayablenotjustastext,butalsoasbibtexentries(andotherformatsastheybecomestandard).Weaimtomakethejournalsiteasand-boxforthecommu-nity.Searchingandfilteringpapers,usingSemanticWebtechnologies,forgeneratingcustompagescontainingthepapersbyaparticularauthororonaparticulartopic.Demonstrationsandontologiespublishedinthejournalwillbemadeavailablelinkingbacktheauthorssiteandarchivingacopyofthesubmission.Hencethemostuptodateversionisvisible,buttheversionatsubmissiontimeisavailable.Ourgoalistomakethisoneofthemostreadablejournalsonyourshelvesandarepositoryofimportantresourcesontheweb.4.ThisissueOurfirstissuepurposelyattemptstoreflectthedi-versityandrangeofresearchshapingWebSemantics,withrepresentativesofmostofourpaperstyles:fivelongresearchpapers,eachviewingthewebinadiffer-entway,andtwoshortpapersdescribinganontologyandademonstratorapplication.KnowledgerepresentationisamajortopicofWebSemantics.Thechallengeistoviewthewebasahugeknowledgebase.In“FromSHIQandRDFtoOWL:TheMakingofaWebOntologyLanguage”HroducetheOWLWebOntologyLanguage,aformallanguageforrepresentingontologiesintheSemanticWeb,recentlyannouncedbyW3Casacandidaterecommendation.OWLoffersthefeaturesculledfromtheresultsofknowledgerepresentationresearch,andwasdesignedontopofRDF.ThepaperdescribeshowOWLwasborn,whatresearchissueshavebeensolved,andwhatremains.AgenttechnologyfacilitatestheuseofWebSeman-tics,andWebServicesarefacilitatedbytheuseofWebSemantics.Thechallengeistoviewthewebasahugeservice-basedmulti-agentsystem.Sycaraetal.in“AutomatedDiscovery,InteractionandComposi-tionofSemanticWebServices”proposeDAML-SfordescribingSemanticWebServices,whichcombinesthewebservicesarchitecturewithSemanticWeb.DAML-SprovidesanabstractdescriptionofWebServices,andcansupportmatchingandinteractionamongwebservices.Thepaperdescribestheimple-mentationoftheDAML-S/UDDIMatchmakerand4Editorial/WebSemantics:Science,ServicesandAgentsontheWorldWideWeb1(2003)15theDAML-SVirtualMachinetoactuallyprototypeSemanticWebServices.DatabasetechnologyisanotherstreamofWebSemantics.SergeyMelniketal.developagenericmodelmanagementsystemcalledRondoin“Devel-opingMetadata-IntensiveapplicationswithRondo”.Thechallengeistoviewthewebastheintegrationofahugenumberofonlineapplications,services,anddatabases.Thesesystemsaretiedtogetherusingmediators,mappings,databaseviews,andtransforma-tionscripts.Modelmanagementreducestheamountofprogrammingneededforintegratingapplications.Thepaperintroduceshigh-leveloperatorstomanip-ulatemodelsandmappingsbetweenmodels,suchaschangepropagation,viewreuse,andreintegration.Semanticannotationisthekeytocreatetheseman-ticcontentoftheSemanticWeb.“Acaseforautomatedlarge-scalesemanticannotation”byDilletal.risestothechallengetoautomaticallycreateahugevolumeoftaggedwebcontentsfromexistingwebpages.ThepaperdescribesSeeker,aplatformforlarge-scaletextanalysis,andSemTag,anapplicationwrittenontheplatformthatperformstheautomatedsemantictag-gingoflargecorpora.Approximately264millionwebpagesaretagged,generatingaround434milliondis-ambiguatedsemanticannotations.Abereretal.s“Startmakingsense:TheChattyWebapproachforglobalsemanticagreements”de-scribesasteptowardsself-learningnetworksofpeersestablishingsemanticoperabilityautomatically.Thechallengeistoharnessthehugenetworkofinter-connecteddatasources,andtocometonegotiatedagreementsonsemantics.Participatingdatasourcesincrementallydevelopglobalagreementinanevo-lutionaryandcompletelydecentralizedprocessthatsolelyreliesonpair-wise,localinteractions.Theauthorsclaimtheirapproachappliestoanysys-temwhichprovidesacommunicationinfrastructure(websitesordatabases,decentralizedsystems,P2Psystems)andofferstheopportunitytostudysemanticinteroperabilityasaglobalphenomenoninanetworkofinformationsharingparties.Ontologyisakeytorealizingsemanticcontents.Golbecketal.havedevelopedanontologybasedontheNationalCancerInstitutesThesaurus.TheneedforacomprehensiveterminologyarosebecausetermswereoftenlocallydevelopedwithinvarioussectionsoftheInstitute.TomaketheknowledgeintheThe-saurusmoreusefulandaccessible,theNationalCancerInstituteandtheUniversityofMarylandhaveworkedtogethertoproduceanOWLontologyfromtheThe-saurus.ToolsareneededtoacceleratetheadvanceofWebSemantics.R.GuhaandRobMcCoolreportonTAP,anexperimentalsystemforidentifyingandresearch-ingdifferenttechnicalissuessuchasscalablequerylanguages,sharingvocabularies,bootstrappingknowl-edgebases,automatedextractionofRDFfromtext,etc.TAPhasbeenusedtocreatelarge-scalesemanticannotationasdescribedbyDillandcolleagues.FutureissueswillincludepapersonSemanticGrid,NaturalL