propertiesbackpage

OpenResearchOnlineTheOpenUniversity'srepositoryofresearchpublicationsandotherresearchoutputsBletchleyParktext:usingmobileandsemanticwebtechnologiestosupportthepost-visituseofonlinemuseumresourcesJournalItemHowtocite:Mulholland,Paul;Collins,TrevorandZdrahal,Zdenek(2005).
BletchleyParktext:usingmobileandsemanticwebtechnologiestosupportthepost-visituseofonlinemuseumresources.
JournalofInteractiveMediainEducation,24(Specia)ForguidanceoncitationsseeFAQs.
c[notrecorded]Version:[notrecorded]Link(s)toarticleonpublisher'swebsite:http://jime.
open.
ac.
uk/2005/24/mulholland-2005-24-t.
htmlCopyrightandMoralRightsforthearticlesonthissiteareretainedbytheindividualauthorsand/orothercopyrightowners.
FormoreinformationonOpenResearchOnline'sdatapolicyonreuseofmaterialspleaseconsultthepoliciespage.
oro.
open.
ac.
ukMulholland,P.
,Collins,T.
&Zdrahal,Z.
(2005).
BletchleyParkText:Usingmobileandsemanticwebtechnologiestosupportthepost-visituseofonlinemuseumresources.
JournalofInteractiveMediainEducation2005(24).
[jime.
open.
ac.
uk/2005/24].
PaulMulholland,TrevorCollins,ZdenekZdrahal,TheOpenUniversity,UK.
Page1Published22ndDecember2005ISSN:1365-893XBletchleyParkText:Usingmobileandsemanticwebtechnologiestosupportthepost-visituseofonlinemuseumresourcesPaulMulholland,TrevorCollins,ZdenekZdrahalAbstract:Anumberoftechnologieshavebeendevelopedtosupportthemuseumvisitor,withtheaimofmakingtheirvisitmoreeducationallyrewardingand/orentertaining.
ExamplesincludePDA-basedpersonalizedtourguidesandvirtualrealityrepresentationsofculturalobjectsorscenes.
Ratherthansupportingtheactualvisit,wedecidedtoemploytechnologytosupportthepost-visitor,thatis,encouragefollow-upactivitiesamongrecentvisitorstoamuseum.
Thisallowedustousethetechnologyinawaythatwouldnotdetractfromtheexistingcuratedexperienceandallowthemuseumtoprovideaccesstoadditionalheritageresourcesthatcannotbepresentedduringthephysicalvisit.
Withinourapplication,calledBletchleyParkText,visitorsexpresstheirinterestsbysendingtext(SMS)messagescontainingsuggestedkeywordsusingtheirownmobilephone.
Thesemanticdescriptionofthearchiveofresourcesisthenusedtoretrieveandorganizeacollectionofcontentintoapersonalizedwebsiteforusewhentheygethome.
Organizationofthecollectionoccursbothbottom-upfromthesemanticdescriptionofeachiteminthecollection,andalsotop-downaccordingtoaformalrepresentationoftheoverallmuseumstory.
Indesigningtheinterfaceweaimedtosupportexplorationacrossthecontentarchiveratherthanjustthesearchandretrievalofspecificresources.
TheservicewasdevelopedfortheBletchleyParkmuseumandhassincebeenlaunchedforusebyallvisitors.
Keywords:SMS,semanticannotation,ontology,museumvisit,collecting,exploration,heritagearchive.
InteractiveDemonstration:Figures8,9and10insection3.
3areavailableasinteractivedemosofBletchleyParkText.
Commentaries:AllJIMEarticlesarepublishedwithlinkstoacommentariesarea,whichincludespartofthearticle'soriginalreviewdebate.
Readersareinvitedtomakeuseofthisresource,andtoaddtheirowncommentaries.
Theauthors,reviewers,andanyoneelsewhohas'subscribed'tothisarticleviathewebsitewillreceivee-mailcopiesofyourpostings.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page2Mulholland,etal.
(2005)1IntroductionInamuseumcontext,newtechnologyisoftenusedeithertoaugment,orprovideasurrogatefor,thephysicalvisit.
ExperimentsinvisitaugmentationincludePDA-basedpersonalizedtourguides(Cheverst,DaviesandMitchell,2002)and"smart"museumobjectsthatcanbephysicallymanipulatedbythevisitor(Ferris,Bannon,Ciolfi,Gallagher,HallandLennon,2004).
Technologiesthatcanbeusedassurrogatesforaphysicalvisitinclude3Dvirtualrealityrepresentationsofgalleriesandothergeographicalareasofculturalinterest(Zára,2004).
Thepost-visitor,whohasrecentlybeentothemuseumhas,however,receivedlessattention.
NotableexceptionsincludeworkattheExploratoriuminSanFrancisco(Fleck,Frid,KindbergandO'Brien-Strain,2002;HsiandFait,2005)andthePeabodyEssexMuseum'sARTscape(Johnson,2004).
TheExploratoriumhaveusedRFIDandPDAtechnologiestoallowavisitortobookmarkanexhibitduringthephysicalvisitandthenlateraccessrelatedresourcesfromthewebsite.
Similarly,ARTscapeallowsavisitortobookmarkexhibitsusingaspeciallyadaptedaudiowand,andthenlaterviewfurtherinformationonthesefromtheirpersonalcollection.
Developingtechnologyforthepost-visitorisattractiveforanumberofreasons.
KravchynaandHastings(2002)inasurveyofvisitorstoamuseumwebsitefoundthat57%ofthemvisitedamuseum'swebsitebothbeforeandafteraphysicalvisittothemuseum.
Despitesuchfindings,museumwebsitesoftenfailtocaterforthepost-visitor,insteadsometimesjustprovidingaccesstosmalldigitalversionsofobjectstheyhavealreadyseenforreal,orinformationthatisnolongerrelevantsuchastraveldetailsandopeningtimes.
Cateringforthepost-visitoralsoallowsthemuseumtoshowcaseheldresourcesthatcannotbeappropriatelydisplayedinthephysicalmuseum.
Thisparticularlyappliestoresourcesthatwouldappealtothosewhohadalreadyvisitedthemuseumandacquiredtheoverallstory,andthatmaybelessmeaningfultoawebsitevisitorwhohasyettoattendthephysicalmuseum.
Anotherreasonfortargetingthepost-visitoristhatencouragingfollow-upactivitiescanbeverybeneficialtothemuseum.
Thepost-visitormaybemotivatedtomakeonlinepurchasesfromthemuseumshop,joininterestgroupsthatpromotethemuseum,ormakeareturnvisit.
Ourapproachthereforeaimstoformulateanswerstothefollowingquestions:Integration:Howcannewtechnologyhelptothreadtogetherthevisitandpost-visitexperienceMotivation:HowcannewtechnologybeusedtoencouragethevisitortoinvestigateaheritagedomainfurtherafterthemuseumvisithasendedExploration:Howcanthepost-visitorbesupportedinexploringanarchiveofcontentassociatedwith,butnotpresented,atthemuseumUsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page3Mulholland,etal.
(2005)OurapproachwasdevelopedandtestedfortheBletchleyParkMuseum,basedinMiltonKeynes,UK.
BletchleyParkwasawartimecodebreakingcentrethatdevelopedtheColossuscomputer.
Intheearly1990s,itbecameaheritagecentrefocusingonthehistoryofcomputingandcodebreaking.
ThemuseumtourdescribesthehistoryofBletchleyPark,itsbuildings,importantcharacters,workandinventions.
Aswellastheexhibitsandothermaterialsmadeavailabletovisitors,BletchleyParkhasanextensivearchivewhichincludeshundredsofhistoricalaccountsandinterviewswithpeopleconnectedtoBletchleyPark.
Thesewereidentifiedastheresourcesthatwouldbemadeavailableafterthevisit.
2Searching,browsingandcollectingOverall,weaimedtoprovideaservicethatwouldallowthevisitortoexploreanarchiveofresourcesbasedoninterestsexpressedduringtheirphysicalvisit.
Thisscenarioproducedanumberofrequirementsthathadtobeborneinmindwhendesigningthetoolforaccessingandusingtheresourcearchive.
First,thepost-visitorwillprobablyhavenopriorknowledgeofthecontentarchive,andwillthereforemostlikelybeinterestedinexploringtherangeofcontentratherthanusingthecontenttoanswerspecificquestionsthattheyalreadyhave.
AsimilarrequirementmotivatedthedevelopmentofthecollectionunderstandingsystemofChang,Leggett,Furuta,Kerne,Williams,BurnsandBias(2004),withwhichanumberofpaintingsbyanartistcanbeorganizedintoacollagetogiveanoverviewoftheirwork.
Second,becauseofthevisitor'slackofpriorknowledgeofthemetadataandclassificationschemesusedbythemuseumtoorganisethecontent,thevisitorneedssufficientscaffoldingforcontentselectionandunderstandinghowdifferentresourcesareconceptuallyrelated.
Othersystemshaveprovideddifferenttypesofscaffoldedinterfacestometadata-annotatedcontent.
Topia(Rutledge,Alberink,Brussee,Pokraev,vanDietenandVeenstra,2003)isahypermediagenerationsystemfororganisingcomponents(e.
g.
textfragmentsorpictures)intoacoherentpresentation.
Thecontentisorganisedintosemanticallysimilarclusterswhicharethenusedtocreateanavigablepresentation.
TopiahasbeentestedwithcontentfromtheRijksmuseuminAmsterdam.
Yee,Swearingen,LiandHearst(2003)developedafacetedbrowsinginterface,wherebytheuserofamuseumwebsitecanselectresourcesrelatedtoparticularmetadatapropertiesbyselectingthemfromdynamicallygeneratedmenusinaconventional-lookingwebinterface.
Similarly,Ontogator(Hyvnen,SaarelaandViljanen,2004)hasbeenusedtoprovideafacetedbrowserinterface,aswellasarecommendationsystem,toaphotographicrepositoryoftheHelsinkiUniversity.
Third,asthepost-visitorhasalreadyexpressedtheirinterestsviaSMS,thewebsiteshouldbeclearlypersonalizedaccordingtotheseinterests.
ThisisconsistentwithFalkandDierking's(2000)modeloffree-choicelearningthatgrewoutoflargescalestudyofmuseumvisitors.
Free-choicelearningisfacilitatedwhenindividualshavecontrolovertheirlearningexperience,includingchoiceoverwhatislearnedandwhen.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page4Mulholland,etal.
(2005)Fourth,fromtheperspectiveofthemuseum,theinterfaceshouldencouragethepost-visitortoexplorefurtherandengagewiththedomain,ratherthanjusttakeaquicklookatthewebsiteandthenmoveon.
HsiandFait(2005)suggesthowpost-visittechnologycouldencouragefurtherlearningbeyondthesinglephysicalvisit.
Neitherstandardsearchorbrowsinginterfacescanbeexpectedtoadequatelymeettheserequirements.
Searchis,ofcourse,personalized,butismoreappropriateforansweringspecificquestionsandfindingspecificresources,ratherthanforopen-endedexploration.
Searchalsorequiresareasonablelevelofdomainknowledgetoformulateappropriatequeries.
Apre-craftedorsimpledatabasedrivenwebsiteforbrowsingthecontentprovidesmorescaffoldingbutisnotverysuitableforprovidingapersonalizedexperience.
Ourapproach,whichwerefertoasSpotlightBrowsing,differsfromtheabove.
Here,thevisitor'sexpressedinterestsareusedtoselectacollectionofresources.
Thiscollectionofresourcesisthenorganizedforthevisitor,givingthemaccesstotheresourcesinthecollectionandalsohighlightingotherrelatedtopicsassociatedwiththecollection.
Thisthereforecontrastswithsearch,wheretheultimategoalistoretrieveasingleresourcethatanswerstheuser'squestion.
Here,theaimistosupporttheexplorationofaresourcecollectionandintroduceadditionaltopicsforfurthercollection-buildingandexploration.
Thiscollection-basedapproachwasmotivatedbyanumberofobservations.
First,theprocessofcollecting,whetherthisbeselectingpapersfromadigitallibrary,gettinginformationfromcolleaguesorcheckingpricesonthestockmarket,iswidelyseentobeanimportantstageincreativityandlearningforwhichnewenablingtechnologiesshouldbedeveloped(Shneiderman,2002).
Second,acollectionofresourcescan"talkback"totheviewer,revealinginterestingpatternssuchascommonalities,contrastsandanomalies.
Thecollectioncanthereforerevealanumberofpropertiesthatperhapscouldnotberealizedfromtheindividualitemsmakingupthecollection.
ThisideadatesbacktotheGestaltPsychologists,suchasWertheimer(1924)andisoftensummarizedinthephrase"Thewholeisgreaterthanthesumoftheparts".
Beliefsastotheentertainmentvalueaswellastheeducationalbenefitsofcollectionscanalsobefoundelsewhere.
Fromthesixteenthtoeighteenthcentury,anumberofEuropeanhouseholdshadaWunderkammerorCabinetofCuriosity(Putnam,2001).
Thesewerepersonalcollectionsofobjectsaccumulatedinacabinetordedicatedroom.
Somecollectionshadanemphasisontheunusualorevenmacabre.
TheWunderkammerservedasaninterestingtalkingpointforvisitorstothehouseholdandwasalsousedasaplacetoruminateandreflectinordertoinspirecreativity.
Anothercharacteristicofcollectionsisthattheycanimposeinterestingchallengesonthecollectorastowhatisputinthecollectionandwhatisleftout.
AwellknownexampleistheTimeCapsule-aboxthatisfilledwithobjectsinordertoexplainatimeand/orplaceforpeopleinthefuturewhomayuncoverit.
Constraintsastotheavailableobjectsandsizeoftheboxmakethisaninterestingchallenge.
SuchactivitieshavealsobeenpresentedinaUsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page5Mulholland,etal.
(2005)museumcontext,forexample,PeterGreenaway's1991exhibitionentitled"100objectstorepresenttheworld"(seePutnam,2001).
Finally,collectionsofresourcesconstituteanarrativeandthereforeofferaperspectivethatcanbedescribedas"true-ish"(Pearce,1995)-neithertruenorfalsebutsignificant.
Byselectingsomeobjectsandorganisingtheminacertainway,thecollectorornarratorisputtingforwardsomeparticularviewpoint,essentiallytellingastorywiththeobjects.
Othersmayorganisethecollectioninadifferentway,orselectdifferentobjectstoorganise.
Differentreadersofthecollectionmayalsointerpretitindifferentways.
Aselectedandstructuredcollectionofobjectsthereforetellsoneofnumberofpossiblestoriesthatcanbepotentiallyreadandinterpretedinanumberofways.
3Ourapproach:Selection-Organization-ExplorationTheprocessofSpotlightBrowsingcanbecharacterizedasprogressingthroughthreesteps:selection,organizationandexploration(seefigure1).
Selectioninvolvesspecifyingtheresourcecollectiontobeexplored.
Organizationinvolvesstructuringthecollectioninorderthatitcanbeeffectivelyused.
Thevisitorcanthenexplorethecollectioninordertopursuetheirinterestsandpossiblyselectresourcesforanewcollection,leadingtofurtherexploration.
OuroverallmodelcanbecomparedwithPetersonandLevene's(2003)trailmodelofnavigationallearning.
Intheirmodel,learningisformulatedasacyclicprocessthatswitchesbetweenstagesofenactingandediting.
Duringtheenactingstage,thelearnerbuildstrails(i.
e.
linearpathwaysthroughlearningresources).
Duringtheeditingstage,thelearnerreflectsupontheenactedtrailsandmodifiesthemasaresult.
Theirmodelofnavigationallearningviewstrailsasbeingfirstclasslearningobjectsintheirownright,aswellastheresourcesthatmakeupeachtrail.
Withinourmodel,acollectionoflearningresourcesisalsoviewedasafirstclasslearningobject.
Oneimportantdifferenceinourmodelisthatacollectionofresourcescanbeorganisedinanumberofdifferentstructures,notonlyintolinearpathways.
Ourcurrenttechnicalimplementationincorporatestheautomatedorganisationofacollectionintoanumberofstructures,includinglinearpathways.
Thisalsodifferentiatesourworkfrompreviouspost-visittechnologies.
Theseexplicitlylinkexhibitstoparticularresources,providingfollow-upmaterialforthatexhibit(e.
g.
HsiandFait,2005;Johnson,2004).
Inourapproach,theoverallcollectionofresourcesrelatedtothechosenSMStermsisorganisedintoanumberofviews.
Theseviewsdonotnecessarilymaintainaone-to-onemappingbetweentheexhibitsandtheadditionalresourcesprovidedforfollow-up.
TherestofthispaperdescribedhowthisapproachhasbeeninstantiatedinthedevelopmentofBletchleyParkTextandsummarisesfindingsfromapreliminaryevaluation.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page6Mulholland,etal.
(2005)1)Selection-Specifyingresourcesofinteresteitherduringorafterthevisit,drawingonsuggestionsprovided2)Organization-Top-downandbottom-upstructuringoftheresourcesusingaknowledgeleveldescriptionoftheresourcesandoverallstoryofthemuseum3)Exploration-BrowsingoftheresourcesarrangedintoanumberofviewssuchaspathwaysandinterlinkedcategoriesFigure1.
ThestepsofSpotlightBrowsing:Selection,OrganizationandExploration.
3.
1Step1:SelectionoftermsviaSMSSelectioninitiallyoccursduringthemuseumvisit.
ThevisitorexpressestheirinterestsbysendingkeywordsviatextmessagewhileatBletchleyPark.
Keywordsrelatetothepeople,places,objects,groupsandworkingrolesassociatedwithBletchleyPark.
Suggestedkeywordscanbefoundonexhibitlabels(seefigure2),buildingsigns,postersondisplayinBletchleyPark,aleafletforvisitors(seefigure3),andmayalsobesuggestedbythetourguide.
Thepostersandleafletsalsoprovideasimplestep-by-stepguidetoBletchleyParkTextandhowitshouldbeused(seefigure4).
Priortothisstageofselection,theprocessofvisitingthemuseum,viewingtheexhibitsandfindingtopicofinterestcanbeviewedasanearlierexplorationstage.
Inthispaperwedonotdescribethispriorexplorationstageasitisnotspecifictoourapproach.
FalkandDierking(2000)provideanextensiveaccountofhowvisitorsexplorephysicalmuseumspaces.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page7Mulholland,etal.
(2005)Figure2.
Anexhibitlabelsuggestingthekeyword"bombe".
BombewasthenameofthedecryptionmachinedevelopedbyAlanTuring.
Intheinstructionsshownontheleafletandposter,thevisitorisadvisedtoselectaroundfivekeywords.
Thisadviceisgiveninordertokeeptheresultingcollectiontoamanageablesizeandalsoencouragethevisitortobeselectiveastowhattheywouldliketoknowmoreabout.
Previousinformaltestsofthesoftwarefoundthataroundfivetermsproduceasufficientlyrichbutstillmanageablecollectionofresources.
Essentially,thisimposesaconstraintonthevisitorrequiringthemtodecidewhichaspectsofBletchleyParktheyseeasmostimportantorinteresting.
Intotal,about150differentkeywordsaredisplayedaroundthesite.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page8Mulholland,etal.
(2005)Figure3.
ThefrontpageoftheBletchleyParkTextvisitorleaflet(left)andthebackpage(right)givingasmallnumberoftheavailablekeywords.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page9Mulholland,etal.
(2005)Figure4.
ThemiddlepagesoftheBletchleyParkTextvisitorleafletgivingstep-by-stepinstructionsonhowtousetheservice.
3.
2Step2:OrganizingthecollectionOncethevisitorgetshome,heorshecanaccesstheircollectionofresourcesbygoingtotheBletchleyParkTextwebsite(http://www.
bletchleypark.
org.
uk/text)andenteringthenumberofthemobilephonetheyusedtosendthemessage.
Acollectionofresourcesrelatedtotheirkeywordsarethenretrieved.
Theretrievalprocessusestheknowledgeleveldescriptionofeachresourceinthearchive.
Anabbreviatedsemanticdescriptionofanhistoricalaccountisshowninpseudo-codeinfigure5.
Eachresourceisrepresentedasastorywhichhasanumberofslots.
Keyslotsinclude"hascentralactor"(i.
e.
thekeypeopleorgroupsfeaturedinthestory),"describesphysicalobject","hastheme"and"describesevent".
Eachoftheseslotscanhavemorethanonevalue.
Thevalueofthe|"describesevent"slotis,likestoryitself,astructurecontainingasetofslots.
Theslotsofaneventvarydependingonitstype(i.
e.
thesubclassofeventtowhichitbelongs).
Twoimportantslots,thatcanbefoundinmanysubclassesofevent,are"hasactor"(i.
e.
thepersonorgroupthatperformstheactivity)and"hasrecipient"(i.
e.
theentityatwhichtheactivityisdirected).
Othercommonslotsincludehas"timespecification"(i.
e.
whentheeventoccurred)and"haslocation".
Returningtofigure5,thisdescribesthestoryashavingAlanTuringasthecentralcharacter,andBombeandEnigmamachineastwophysicalobjectsthatfeatureinthestory.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page10Mulholland,etal.
(2005)Thestorycontainstwoevents.
ThefirstisconcernedwithAlanTuring'sdevelopmentoftheBombemachine.
Thesecondisconcernedwithhisdecryptionoftheenigmamachine.
Thecollectionofresourcesisretrievedaccordingtothekeywordsinthetextmessage.
Eachkeywordmapstoaconceptusedforthesemanticdescriptionofthestories.
AvariantoftheSoundexphoneticalgorithmisalsousedtohelpdetectmistypedkeywordsinthetextmessage.
Thecollectionismadeupofresourcesthatmatchanyoftheconceptsassociatedwiththechosenkeywordsinthetextmessage.
Onceretrieved,theresourcesareorganizedforpresentation.
Twostructuresintowhichtheresourcesareorganizedarepathwaysandcategories.
Shipman,Furuta,Brenner,ChungandHsieh(1998)havedemonstratedtheeducationalbenefitsoforganizingwebpagesintolinearpaths.
TheirWalden'sPathsenvironmentcanbeusedbyateachertoorganizeasetofwebpagesintoapathforusebystudents.
Thepathmayexpresssomemessageacrossthecollectedcontent.
Story1Hascentralactor:AlanTuringDescribesphysicalobject:Bombe,EnigmamachineHastheme:Code-breakingEvent1(oftypedeveloping)Hasactor:AlanTuringHasrecipient:BombeEvent2(oftypedecrypting)Hasactor:AlanTuringHasrecipient:EnigmamachineFigure5.
Anabbreviatedsemanticdescriptionofanhistoricalaccountshowninpseudo-code.
Inoursystem,pathwaysareconstructedautomaticallyusingtheknowledgeleveldescriptionoftheeventsdescribedinthestories.
Figure6showsapathwayfromBombetoJoanSmithinvolvingtwostories.
ThestorycontainingEvent1canbeusedasastartingpointforapathwayfromBombetoJoanSmithasithasBombeasavalueofoneofitsslots.
ThisstepinthepathwaytakesustoAlanTuring.
TherelationshipbetweenBombeandAlanTuringisgeneratedbasedontheeventtypeandtheparticularslotsofwhichBombeandAlanTuringarevalues.
Forexample,astheeventtypeis"developing"andAlanTuringistheactorandBombeistherecipient,thisistranslatedas"BombewasdevelopedbyAlanTuring".
AsecondstorycontainingEvent3isthenusedtolinkAlanTuringtoJoanSmith.
AsAlanTuringandJoanSmitharebothactorsinthisevent,andtheeventtypeis"meeting",thisstepinthepathwayistranslatedas"AlanTuringmetwithJoanSmith".
Inthecurrentimplementation,theshortestpathbetweentwoconcepts(e.
g.
peopleUsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page11Mulholland,etal.
(2005)orplaces)isgenerated.
Othermethodsofpathselectionareunderdevelopment.
Currently,apathbetweentwoconceptscaninvolveuptoeightseparatestories.
Theweb-basedpresentationusingthepathstructureisdescribedlaterandshowninfigure8.
Story1containing:Event1(oftypedeveloping)actor:AlanTuringrecipient:BombeStory2containing:Event3(oftypemeeting)actor:AlanTuring,JoanSmithBombewasdevelopedbyAlanTuring.
AlanTuringmetJoanSmith.
StorypathKnowledgeleveldescriptionoftheeventsFigure6.
Astorypathwayandthesemanticdescriptionsofitscomponents.
AnothermethodfororganizingresourcesusedwithinBletchleyParkTextistostructuretheresourcesintoasetofinterlinkedcategories.
Previousworkhasshownthatorganizingasetofresourcesintocategoriesfacilitatestheirexploration.
Inalongitudinalstudyofsearchengineuse,Kki(2005)foundthatwhenauser'sgoalismoreexploratory,andlessdirected,organizingtheresultsofasearchintocategoriesfacilitatestheiruse.
Categorieswerehowever,notbeneficialwhentheuserwastryingtofindspecificresourcesandanswerspecificquestions.
Thisfindingalignswellwithourdecisiontousecategoriesinordertosupporttheexplorationofacollectionofresources.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page12Mulholland,etal.
(2005)Slot:ValueS1S2S3S4S5TotalActor:AlanTuring3Location:Hut83Recipient:Bombe2ActivitiesofAlanTuringS1,S2,S3EventsinHut8S3,S4,S5UsingBombeS2,S4Hut8washeadedbyAlanTuringAlanTuringdevelopedtheBombeAlanTuringwasheadofHut8BombewasdevelopedbyAlanTuringFromRelationToAlanTuringheadedHut8Hut8washeadedbyAlanTuringAlanTuringdevelopedBombeBombewasdevelopedbyAlanTuringHut8workedwithHut4GordonWelchmanwasheadofHut6Top-downinterlinkingofcategoriesBottom-upcategorizationofresourcesInterconnectedcategoriesFactsfromthetourstoryPropertiesofthestorycollectionFigure7.
Organizingacollectionofresourcesintoasetofinterlinkedcategories.
Theprocessofgeneratinginterlinkedcategoriesisillustratedinfigure7.
Inthisscaleddownexample,fivestories(S1toS5)areorganizedintothreecategories,twocontainingthreestoriesandonecontainingtwostories.
Categoriescontainresourcesthathaveslotsandvaluesincommon.
Intheexample,thethreestoriescontainingtheslot-valuepairhasactor:AlanTuringformagroup,asdothethreestoriescontaininghaslocation:Hut8.
Forwebpresentation,theslotnamesaretranslatedintoamorereadableform.
Hasactorbecomes"Activitiesof"andhaslocationbecomes"Eventsin".
Toprovideadditionalscaffolding,andexplainwhythesecategoriesmayco-occurinacollection,thecategoriesareinterlinkedusingaformalrepresentationofthetour.
Forthis,thetourstorywasdecomposedintoasetoffacts,eachmakingarelationshipbetweentwoconcepts,suchas"AlanTuringwasheadofHut8"and"AlanTuringdevelopedtheBombe".
TheseareusedtolinkthecategoriesAlanTuringandHut8.
Therelationshipisexpressedas"AlanTuringUsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page13Mulholland,etal.
(2005)washeadofHut8"orconversely,"Hut8washeadedbyAlanTuring".
Inthisorganisation,resourcescanfeatureinmorethanonecategory,forexample,S3featuresinallthreecategories.
Analternativeorganisation,alsofeaturedinBletchleyParkText,organisestheresourceshierarchicallyinorderthateachresourcefeaturesinonlyonecategory.
Thestorycategoriesarethereforecreatedbottom-upfromthecollection,butthelinksbetweenthecategoriesarecreatedtop-downfromthemuseumtour.
Theweb-basedpresentationusingtheaboveinterlinkedcategorystructureisdescribedlaterandshowninfigure9.
3.
3Step3:ExploringthecollectionOnceretrievedandorganized,thecollectionofresourcesispresentedforexplorationbythevisitor.
Anumberofdifferentviewsonthecollectionareprovided.
OneoptionallowsthevisitortomakepathwaysthroughtheresourcecollectionthatconnectconceptsmentionedintheirSMSmessage(seefigure8).
Inthisexample,thevisitorhaschosenfiveconcepts:AlanTuring,StonyStratford,BlockFandHeathRobinsonMachine.
Usingthepull-downmenuonthelefthandside,thevisitorselectswhichofthefiveconceptstheywouldliketoconnect.
Apathwayconnectingthetwoconceptsisthengenerated.
Infigure8,apathwayconnectingAlanTuringtoStonyStratfordisshown.
Thisrequiredtwostories.
ThefirststoryconnectsAlanTuringtoHut7.
ThesecondstoryconnectsHut7toStonyStratford.
Thevisitorcanreadoneofthestoriesinfullbyclickingonitstitle.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page14Mulholland,etal.
(2005)Figure8.
MakingpathwaysthroughthestoriesbyselectingconceptsfromtheSMSmessage.
Aninteractiveversionofthisfigureisavailableathttp://cipherweb.
open.
ac.
uk/connections.
php.
Afurtheroptionorganizesthecollectionintoanumberofinterlinkedcategories(seefigure9).
Thecategoriesareshownontheleft.
Totherightofthisarethesummariesofthestoriesthatbelongtotheselectedcategory.
Theshadedboxshowntothetop-rightshowslinksbetweenthiscategoryandotheravailablecategorieswithinthecollection.
Inthefigure,theselectedcategoryis"EventsinBlockF".
Thishaslinkstotwoothercategorieswithinthecollection.
TheselinksaretoColossus(thatwasusedinBlockF)andtheNewmanry(agroupaworkersbasedinBlockF).
Ascanbeseen,manyofthesecategoriesfeatureconceptsthatwerenotinthevisitor'stextmessage(e.
g.
theNewmanry).
However,theseconceptsco-occurinthecollectionofstoriesUsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page15Mulholland,etal.
(2005)abouttheirchosentopics.
Theyarethereforeofpossibleinteresttothevisitor,anditisalsolikelythat,ingainingaknowledgeoftheirchosenterms,thevisitorwilldevelopsomeappreciationoftheseadditionalrelatedconcepts.
Figure9.
Thecollectionofstoriesorganizedintointerlinkedcategories.
Aninteractiveversionofthisfigureisavailableathttp://cipherweb.
open.
ac.
uk/categories.
php.
Anotherversionoftheinterlinkedcategoryviewseparatesoutcategoriesastowhetherornottheyaredirectlyrelatedtotheconceptscontainedinthevisitor'sSMSmessage(seefigure10).
Thisisrepresentedusingatorchmetaphor.
Categoriesdirectlyrelatedtothetextmessageareshowninthebrightareaofthetorchlight.
Forexample,thecategory"ActivitiesofAlanTuring"isincludedhereasAlanTuringwasoneofthetopicsincludedintheirtextmessage.
Themostcommonlyoccurringcategoriesnotdirectlyrelatedtothetextmessageareshownunderneath,attheedgeofthetorchlight.
ThisviewwasdesignedinUsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page16Mulholland,etal.
(2005)ordertodrawthevisitor'sattentiontotopicsinwhichtheydidnotexpressaninterestbutwhicharecloselyassociatedwiththeconceptsintheirtextmessage.
Thisisintendedtoencouragethevisitortoexplorefurther,beyondtheconfinesoftheirspecifiedcollection,or,followingthemetaphor,encouragethemtomovethetorchlighttoanotherareaofthearchive.
Onceagainthevisitorcanreadanystorybyclickingonitstitle.
Linksbetweentheselectedcategoryandothercategoriesareshownintheshadedboxdisplayedtothetoprightofthescreen.
Figure10.
Thecollectionorganizedintocategoriesdirectlyandindirectlyrelatedtothekeywordsinthetextmessage.
Aninteractiveversionofthisfigureisavailableathttp://cipherweb.
open.
ac.
uk/spotlight.
php.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page17Mulholland,etal.
(2005)3.
4Step1revisited:SelectingnewtermsviathewebsiteThefinaloptionavailabletothevisitorallowsthemtomodifytheirselectedterms(seefigure11).
Theboxtothetopleftshowstheoriginallyselectedtopicsinaeditabletextarea.
Thevisitorcandeleteanyofthesethattheywishtoremovefromtheircollection.
Justbelowthis,topicsaresuggestedthatstatisticallyco-occurwiththechosentopics.
Uptosixoftheseareshown.
Thislistofsuggestionsisthereforegeneratedbottom-upfromthecurrentcollection.
Furtherdownthelefthandsideofthescreenarenewtopicsrelatedtotheoriginaltopicsbythefactrepresentationofthetour.
Forexample,thetopic"Bombe"issuggestedhereasAlanTuringdevelopedtheBombe.
Thissecondlistofsuggestionsifthereforegeneratedtop-downfromthetourfacts.
Thereisoftensomeoverlapbetweenthetopicssuggestedtop-downandthosesuggestedbottom-up.
Infigure11,MaxNewmanfeaturesonbothlistsofsuggestions.
Heisincludedonthefirstlistbecausehefeaturedinalargenumberofstorieswithinthecurrentcollection.
HeisincludedonthesecondlistbecauseatourfactlinkshimtoAlanTuringandAlanTuringwasincludedasatopicintheoriginalmessage.
Anydifferencesbetweenthesuggestedtopicsgeneratedtop-downandbottom-uparelefttothevisitortoponder.
Thecurrentsystemdoesnotresolvethesedifferencesfortheuser.
Ifthevisitorwishestomoveconceptuallyfurtherawayfromtheircurrentcollection,alargerlistofapproximately650topicsispresentedontherightofthescreen.
Thisisshowninordertogivethevisitoranappreciationofthescaleoftheoverallcontent.
Itisalsoprovidedtoprovokecuriosityandmakethemwonderwhatthesetopicsmeanandwhatresourcesmayberevealediftheyareselected.
Consistentwiththeinstructionsprovidedduringthephysicalvisit,thevisitorisadvisedtoselectaboutfivetopicsandisallowedamaximumofnine.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page18Mulholland,etal.
(2005)Figure11.
Choosinganewsetofkeywords.
FromaHCIperspective,theinterfaceforselectinganewcollectionhassomeunusualfeatures:Constrainedchoices:Thevisitorcanonlyselectfivetopicsbutispresentedwithover650tochoosefrom.
Uninformeddecisionmaking:Thevisitormaynotknowwhethertheywouldbeinterestedinreadingaboutsomeunknowntopic,buttheonlywaytofindoutistoselectit.
Unresolvedinconsistencies:Differencesbetweenthetop-downandbottom-upgeneratedsuggestionsareleftunresolvedforthevisitortoconsider.
Thesefeaturesalsopartlycharacterizetheinitialselectionoftopicsduringthephysicalvisit.
ThesecharacteristicsdonothelptheuserretrieveefficientlyinformationfromtheUsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page19Mulholland,etal.
(2005)archive,howeverthatwasneverthegoal.
Efficientretrievalandtheselectionofspecificresourcesisservedbysearchengines.
Theaimherewastoprovideresourcesofinterestbutalsoprovokecuriosityandencouragefurtherengagementwiththecontent.
Theabovecharacteristicsseempromisingwhenthegoaliscuriosityratherthanefficiency.
Gaver,BeaverandBenford(2003)arguethecasefordesigningambiguityintohumancomputerinterfacesinordertomotivateengagementandactiveinterpretationbytheuser.
Theyarguethatalthoughambiguitycanbefrustratingitcanalsobe"intriguing,mysteriousanddelightful".
Animportanttrade-offinsuchinterfacesistoensurethatthefrustrationlevelisnottoohigh,andispaidoffbythepositivequalitiesofthisformofinteraction.
4PreliminaryevaluationApreliminaryevaluationofBletchleyParkText,describedinmoredetailinMulholland,CollinsandZdrahal(2005),wascarriedoutinparticipationwithagroupof35pupilsfromalocalsecondary(post11)school.
AshortpresentationaboutBletchleyParkTextwasgivenatthebeginningoftheirvisitandeachwasgivenaleafletcontainingabout150possibletextterms,organisedintopeople,placesandobjects.
ThepupilsthenspentaroundthreehoursexploringBletchleyParkwiththehelpofatourguide.
Duringthevisit,allwereabletosendtextmessagesand20ofthepupilsaccessedthewebsiteintheirfreetime.
AvolunteergroupofsixpupilsusedBletchleyParkTexttohelpthemwriteanessay.
ThetopicsoftheessaysincludedtheideasofAlanTuring,theBombedecryptionmachine,andmethodsusedforthedecryptionofnavalcommunications.
Ananalysisoftheseessaysshowedthattheyincludedconceptsthatwerenotpresentedduringthephysicalvisitanddrewonmultiplestoriesfromthearchive.
Overall,thestudyshowedthatBletchleyParkTextcouldbeused,thatthepupilsfoundthismethodofaccessingandusingthecontentenjoyable,andthatitplayedapartindevelopingacontinuinginterestinthetopic.
Forexample,oneofthesixvolunteerpupilsaskedforreassurancethathisaccountonBletchleyParkTextwouldbepreservedsothathecouldcontinuetouseit,andanothertwohasacquiredbooksrelatedtotheiressaytopicforfurtherfollow-upintheirfreetime.
Furtherstudieswillcomparetheapproachagainstmoretraditionalsearchingandbrowsinginterfacesandtherelativeroleplayedbyeachoftheviewsonthecollectionprovidedintheinterface.
5ConclusionsBletchleyParkTextisaninstantiationofanovelapproachtotheexplorationofanarchiveofresources,calledSpotlightBrowsing.
Thisallowstheusertospecifyacollectionofinterest,haveitautomaticallyorganizedintopathwaysandinterconnectedcategories,explorethecollection,andthenredefinethecollectioninordertoexplorefurther.
BletchleyParkTextwaslaunchedasavisitorserviceinMay2005.
Futureworkwillinvolveestablishingthescalabilityandgeneralityoftheapproach,todeterminewhether,andunderwhatconditions,SpotlightBrowsingcansupporttheexplorationofothercontentarchives.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page20Mulholland,etal.
(2005)6AcknowledgementsThisworkhasbeensupportedbytheEuropeanUnionInformationSocietyTechnology(IST)fundedCIPHERproject(IST2002-04-22)andtheOpenUniversityfundedSUbTLEproject.
7ReferencesChang,M.
,Leggett,J.
,Furuta,R.
,Kerne,A.
,Williams,J.
,Burns,S.
,andBias,R.
(2004).
Collectionunderstanding.
ProceedingsoftheJointConferenceonDigitalLibraries(JCDL'04),Tucson,Arizona.
Cheverst,K.
,Davies,N.
,andMitchell,K.
(2002).
TheRoleofAdaptiveHypermediawithinaContext-AwareTouristGUIDE.
CommunicationsoftheACM,45(5),47-51.
Falk,J.
H.
andDierking,L.
D.
(2000).
Learningfrommuseums:Visitorexperiencesandthemakingofmeaning.
RowmanandLittlefield:WalnutCreek,CA.
Ferris,K.
,Bannon,L.
,Ciolfi,L.
,Gallagher,P.
,Hall,T.
,andLennon,M.
(2004).
ShapingExperiencesintheHuntMuseum:ADesignCaseStudy.
ProceedingsoftheACMConferenceonDesigningInteractiveSystems(DIS'04),Cambridge,MA.
Fleck,M.
,Frid,M.
,Kindberg,T.
andO'Brien-Strain,E.
(2002).
Rememberer:Atoolforcapturingmuseumvisits.
TechnicalreportHPL-2002-187.
HPLabs.
Gaver,W.
,Beaver,J.
,andBenford,S.
(2003).
Ambiguityasaresourcefordesign.
ProceedingsoftheACMSIGCHIconferenceonHumanfactorsincomputingsystems(CHI'03),Ft.
Lauderdale,Florida.
Hsi,S.
andFait,H.
(2005).
RFIDenhancesvisitors'museumexperienceattheExploratorium.
CommunicationsoftheACM,48(9),60-65.
Hyvnen,E.
,Saarela,S.
andViljanen,K.
(2004).
Applicationofontologytechniquestoview-basedsemanticsearchandbrowsing.
ProceedingsoftheFirstEuropeanSemanticWebSymposium(ESWS'04),Heraklion,Greece.
Johnson,B.
(2004).
Beyondon-linecollections:Puttingobjectstowork.
MuseumsandtheWeb2004,WashingtonDC.
Kravchyna,V.
andHastings,S.
K.
(2002).
InformationalValueofMuseumWebSites.
FirstMonday,7(2).
URL:http://firstmonday.
org/issues/issue7_2/kravchyna/index.
html.
Kki,M.
(2005).
Findex:Searchresultcategorieshelpuserswhendocumentrankingfails.
InProceedingsoftheACMSIGCHIconferenceonHumanfactorsincomputingsystems(CHI'05),Portland,Oregon.
Mulholland,P.
CollinsT.
,andZdrahal,Z.
(2005).
Spotlightbrowsingofresourcearchives.
ProceedingsoftheACMConferenceonHypertextandHypermedia(HT'05),Salzburg,Austria.
Pearce,S.
M.
(1995).
Oncollecting:AninvestigationintocollectingintheEuropeantradition.
Routledge:London.
Peterson,D.
andLevene,M.
(2003).
Trailrecordsandnavigationallearning.
LondonReviewofEducation,1,(2003),207-216.
UsingmobileandsemanticwebtechnologiesJournalofInteractiveMediainEducation,2005(number)Page21Mulholland,etal.
(2005)Putnam,J.
(2001).
ArtandArtifact:TheMuseumasMedium.
Thames&Hudson:London,UK.
Rutledge,L.
,Alberink,M.
,Brussee,R.
,Pokraev,S.
,vanDieten,W.
andVeenstra,M.
(2003).
Findingthestory-Broaderapplicabilityofsemanticsanddiscourseforhypermediageneration.
ProceedingsoftheACMConferenceonHypertextandHypermedia(HT'03),Nottingham,UK.
Shipman,F.
,Furuta,R.
,Brenner,D.
,Chung,C.
andHsieh,H.
(1998).
Usingpathsintheclassroom:experiencesandadaptations.
ProceedingsoftheACMConferenceonHypertextandHypermedia(HT'98),Pittsburgh,Pennsylvania.
Shneiderman,B.
(2002).
Leonardo'sLaptop:HumanNeedsandNewComputingTechnologies.
MITPress:Cambridge,MA.
Wertheimer,M.
(1924).
Gestalttheory.
EnglishtranslationinEllis,W.
D.
(Ed.
)(1939).
ASourceBookofGestaltPsychology.
HarcourtBrace:NewYork.
AvailablefromtheGestaltarchive,URL:http://gestalttheory.
net/archive.
Yee,K.
-P.
,Swearingen,K.
,Li,K.
andHearst,M.
(2003).
Facetedmetadataforimagesearchandbrowsing.
ProceedingsoftheACMSIGCHIconferenceonHumanfactorsincomputingsystems(CHI'03),Ft.
Lauderdale,Florida.
Zára,J.
(2004).
VirtualRealityandCulturalHeritageontheWeb.
Proceedingsofthe7thInternationalConferenceonComputerGraphicsandArtificialIntelligence,Limoges,France.