PersonalizationServicesfore-Learningin
theSemanticWeb
NicolaHenzea,1a
ISI–SemanticWebGroup,
UniversityofHannover&ResearchCenterL3S
Abstract.ThePersonalReaderframeworkimplementsaservice-basedarchitec-turefordevelopingandmaintainingpersonalizationfunctionalitiesontheSemanticWeb,stemmingfromdisciplineslikee.g.adaptivehypermediasystemsorcollabo-rativefilteringsystems.Amodularframeworkofcomponents/services-forpro-vidingtheuserinterface,formediatingbetweenuserrequestsandavailableperson-alizationservices,forusermodeling,forprovidingpersonalrecommendationsandcontextinformation,etcetera,isthecoreofthePersonalReaderframework.WhenauserisviewingsomeWebContent(the\"Reader\"partofthePersonalReader)s/hereceivesadditional,personalinformationonthecontextofthisparticularWebcontent(the\"Personal\"partofthePersonalReader).PersonalReadershavebeendevelopedfortheareaofe-Learning(Java,SemanticWeb),andforbrowsingsci-entificpublications.
Keywords.PersonalizationServices,PersonalizationArchitectures,SemanticWeb
1.Introduction
WiththeideaofaSemanticWeb[2]inwhichmachinescanunderstand,processandreasonaboutresourcestoprovidebetterandmorecomfortablesupportforhumansininteractingwiththeWorldWideWeb,thequestionofpersonalizingtheinteractionwithwebcontentisathand.Intheareaofadaptivehypermedia,researchhasbeencarriedouttounderstandhowpersonalizationandadaptationstrategiescanbesuccessfullyap-pliedinhypertextsystemsandhypertextlikeenvironments.Ithasbeenstatedthatintheareaofadaptivehypermediaandofadaptiveweb–basedsystems,thefocusofdevelopedsystemshasbeensofaronclosedworldsettings.Thismeansthatthesesystemsworkonafixedsetofresourceswhicharenormallyknowntothesystemdesignersatdesigntime(seethediscussiononclosedcorpusadaptivehypermedia[4]).Thisobservationalsorelatestothefactthattheissueofauthoringadaptivehypermediasystemsisstilloneofthemostimportantresearchquestionsinthisarea,seee.g.[3].AgeneralizationofadaptivehypermediatoanAdaptiveWebdependsthereforeonasolutionoftheclosedcorpusprobleminadaptivehypermedia.WithinthePersonalReaderproject,weproposeanarchitectureforapplyingsomeofthetechniquesdevelopedinadaptivehypermediatoanopencorpus.Amodularframeworkofcomponents/services-forprovidingtheuser
to:NicolaHenze,ISI-SemanticWebGroup,UniversityofHannover&ResearchCenter
L3S,Appelstr.4,D-30167HannoverTel.:+4951176219716;Fax:+4951176219712;E-mail:henze@l3s.de
1Correspondence
Figure1.ScreenshotofthePersonalReaderforlearningabouttheSemanticWeb.ThePersonalReadercon-sistsofabrowserforlearningresourcesthereaderpart,andaside-barorremote,whichdisplaystheresultsofthepersonalizationservices,e.g.individualrecommendationsforlearningresources,contextualinformation,pointerstofurtherlearningresources,quizzes,examples,etc.thepersonalpart.
interface,formediatingbetweenuserrequestsandavailablepersonalizationservices,forusermodeling,forprovidingpersonalrecommendationsandcontextinformation,etcetera,isthecoreofthePersonalReaderframework[7].Thecommunicationsbetweenallcomponents/servicesissyntacticallybasedonRDFdescriptions.E.g.therequestforgettingpersonalrecommendationsforalearningresourceforacertainuserisprovidedbyanRDFdescriptionwhichisexchangedbetweenthecomponentsmediatorandper-sonalrecommendations.Thuseachcomponentisaservice,whichisusuallyindependentfromtheothersandwhichcaninteractwiththemby\"understanding\"theRDFnotifica-tionstheysend.Thecommon\"understanding\"isrealizedbyreferringtosemanticsintheontologiesusedintheRDFdescriptionswhichprovidethevalidvocabulary(see[6,7]).PrototypesofPersonalReadershavebeendevelopedfortheareaofe-Learning(Java,SemanticWeb),andforbrowsingscientificpublications.
2.Proof-of-Concept:PersonalReadersfore-LearningandforBrowsingScientificPublications2.1.PersonalReadersfore-Learning
ThePersonalReadersfore-Learning[5](seeFigure1)providealearnerwithapersonalinterfaceforregardinglearningresources:thePersonalAnnotationServicerecommendsthelearnernextlearningstepstotake,pointstoexamples,summarypages,moredetailedinformation,etc.,andalwaysrecommendsthemostappropriateoftheseinformationac-cordingtothelearner’scurrentknowledge,his/herlearningstyle,learninggoal,back-ground,etc.ThePersonalsearchserviceextractsinformationfromtheactuallyregardedlearningresourceandchecksforrelatedinformationinothere-Learningcorpora,andrecommendsretrievedresults.IfyouwanttosetupyourownPersonalReaderinstance
Figure2.ScreenshotofthePersonalPublicationReader.Whenauserisviewingsomepublication,s/here-ceivesadditional,personalinformationonthecontextofthispublicationwithintheREWERSEproject:back-groundinformationaboutthepersonsandworkinggroupscarryingoutthiskindoforrelatedresearch,addi-tionalinformationabouttheauthors,etc.
foracourseyouarerunning,youneedtoprovideRDFdescriptiononthelearningre-sourcesofthiscourse(examplesofsuchRDFdescriptionscanbefoundfollowingthelinkResourcesonthisprojectpage,andalinktosomedomainontologydescribingtheapplicationdomainofyourcourse,whichyoualsousetoannotateyourresources.Highlights:
•easycreationofpersonalizedReadersforlearningobjectsannotatedaccordingtoLOMstandard;
•demonstrates:re-usablepersonalizationfunctionalityfore-Learningcourses;•reasoningforthepersonalizationservicesisrealizedusingTRIPLE[9]2.2.ThePersonalPublicationReader
ThePersonalPublicationReader[1](seeFigure2)hasbeendevelopedfortheNetworkofExcellenceREWERSEforprovidingapersonalinterfacetothepublicationsdevelopedintheproject:Allweb-pagescontaininginformationaboutpublicationsoftheREW-ERSEnetworkareperiodicallycrawledandnewinformationisautomaticallydetected,extractedandindexedintherepositoryofsemanticdescriptionsoftheREWERSEnet-work.Thisinformation,withextractedinformationontheprojectREWERSE,onpeopleinvolvedintheproject,theirresearchinterests,etc.,isusedtoprovidemoreinforma-tiononeachpublication:whohasauthoredit,whichresearchgroupsarerelatedtothiskindofresearch,whichotherpublicationsarepublishedbytheresearchgrouporbythisauthor,whichotherpublicationsareonthesimilarresearch,etc.Highlights:
•automatizedannotationofWebdata:automaticextractionofWebdata,andau-tomatizedannotationofextracteddatawithmeaningfulsemanticinformation(poweredbytheLixtoSuite,www.lixto.com);
•demonstrates:personalizedcontentsyndication;
•reasoningforthepersonalizationserviceisrealizedJena’sRDQLlanguage[8].3.Conclusion
Wehavepresentedaframeworkfordesigning,implementingandmaintainingadaptiveReaderapplicationsfortheSemanticWeb.ThePersonalReaderframeworkisbasedontheideaofestablishingpersonalizationfunctionalityasservicesontheSemanticWeb.TherealizationofpersonalizationfunctionalityisdoneonthelogiclayeroftheSemanticWebtower,makinguseofdescriptionandrulelanguagerecentlydevelopedinthecontextoftheSemanticWeb.Wehavetestedtheframeworkwithexamplereadersintheareaofe-Learning(Javaprogramming,SemanticWeb),andforbrowsingscientificpublicationsoftheREWERSEproject.Thecurrentstateoftheprojectcanbefollowedatwww.personal-reader.de,wherealltherealizedprototypesareavailable,too.
Acknowledgments
ThisworkhaspartiallybeensupportedbytheEuropeanNetworkofExcellenceREWERSE-ReasoningontheWebwithRulesandSemantics.WewouldliketothankFabianAbel,RobertBaumgartner,TobiasBuchloh,StefanDecker,PeterDolog,LiliaCheniti-Belcadhi,ChristianEnzi,MarcHerrlich,DennisKohlmetz,MatthiasKriesell,KashifMushtaq,WolfgangNejdl,MichaelSintek,SaschaTönnies,andKaiTomaschewskifortheirsupportingettingtheideaofPersonalReadersintoreality.
References
[1]RobertBaumgartner,NicolaHenze,andMarcusHerzog.ThePersonalPublicationReader:
IllustratingWebDataExtraction,PersonalizationandReasoningfortheSemanticWeb.InEuropeanSemanticWebConferenceESWC2005,Heraklion,Greece,May29-June12005.[2]TimBerners-Lee,JimHendler,andOraLassila.TheSemanticWeb.ScientificAmerican,May
2001.
[3]P.DeBra,A.Aerts,D.Smits,andN.Stash.AHA!version2.0:Moreadaptationflexibilityfor
authors.InProceedingsoftheAACEELearn’2002conference,October2002.
[4]PeterBrusilovsky.AdaptiveHypermedia.UserModelingandUser-AdaptedInteraction,
11:87–110,2001.
[5]NicolaHenze.PersonalReaders:PersonalizedLearningObjectReadersfortheSemanticWeb.
In12thInternationalConferenceonArtificialIntelligenceinEducation,AIED’05,Amster-dam,TheNetherlands,July2005.
[6]NicolaHenze,PeterDolog,andWolfgangNejdl.Reasoningandontologiesforpersonalized
e-learning.EducationalTechnology&Society,7(4),2004.
[7]NicolaHenzeandMatthiasKriesell.PersonalizationFunctionalityfortheSemanticWeb:
ArchitecturalOutlineandFirstSampleImplementation.In1stInternationalWorkshoponEngineeringtheAdaptiveWeb(EAW2004),Eindhoven,TheNetherlands,2004.[8]RDQL-querylanguageforRDF,Jena,2005.http://jena.sourceforge.net/RDQL/.
[9]MichaelSintekandStefanDecker.TRIPLE-anRDFQuery,Inference,andTransformation
Language.InInternationalSemanticWebConference(ISWC),Sardinia,Italy,2002.
