d 12.6 final testing report - ariadne pluslegacy.ariadne-infrastructure.eu › ... ›...
TRANSCRIPT
D 12.6 – Final Testing Report
Version 1.0 (final)
January 26th, 2016
Grant Agreement number:
313193
Project acronym: ARIADNE
Project title: Advanced Research Infrastructure for Archaeological Dataset Networking in Europe
Funding Scheme: FP7-INFRASTRUCTURES-2012-1
Project co-ordinator name, Title and Organisation:
Prof Franco Niccolucci, PIN Scrl - Polo Universitario "Città di Prato"
Tel: +39 0574 602578
E-mail: [email protected]
Project website address: www.ariadne-infrastructure.eu
ARIADNED12.6FinalTestingReport
2
The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7-INFRASTRUCTURES-2012-1) under grant agreement n° 313193.
Authors: Anthony Corns
Evie Monaghan
Louise Kennedy
Marie Brohet
Contributing partners
Quality control review
The Discovery Programme
Holly Wright, UoY ADS
Paola Ronzino, PIN
Franco Niccolucci, PIN
ARIADNED12.6FinalTestingReport
3
Table of Contents ExecutiveSummary...........................................................................................................4
1 IntroductionandObjectives.........................................................................................5
2 UseRequirements,ServiceDesignandImplementation...............................................7
3 InfrastructureDesign...................................................................................................83.1 InfrastructureElements.............................................................................................................83.2 ContentProvisionWorkflow.....................................................................................................9
4 Usability&Testing.....................................................................................................11
4.1 ResultsofD12.4.......................................................................................................................114.2 TestingMethodologies............................................................................................................154.3 InternalTesting.......................................................................................................................17
4.3.1 Methodology..................................................................................................................................174.3.2 Testing............................................................................................................................................17
4.4 OnlineSurvey..........................................................................................................................234.4.1 SystemUsabilityScale(SUS)...........................................................................................................234.4.2 SurveyResults.................................................................................................................................26
5 Conclusions................................................................................................................355.1 Registrytooltesting................................................................................................................355.2 MOReaggregationtooltesting...............................................................................................365.3 Portaltesting...........................................................................................................................36
6 References.................................................................................................................37
7 Annex1:OnlineSurvey..............................................................................................39
ARIADNED12.6FinalTestingReport
4
ExecutiveSummary
This document is a deliverable (D12.6 Final Testing Report) of the ARIADNE project (“AdvancedResearch Infrastructure forArchaeologicalDatasetNetworking in Europe”),which is fundedunderthe European Community's Seventh Framework Programme. D12.6 is associated with Task 12.4Testing within WP12, which is titled Implementing Interoperability, and falls within the largerARIADNEconceptualframeworkfortheARIADNEe-Infrastructure.It isanextensionofD12.4Initialinfrastructure testing report, and reports the outcomes of the final acceptance testing, taking onboard the outcomes of the initial testing. This deliverable was also used to inform Tasks 12.3Implementing integration and 12.4 Testing, to facilitate fine tuning of the infrastructure beforecompletionoftheproject.
As outlined in D12.3 one of the goals of the ARIADNE infrastructure is to integrate data andmetadata from different providers across Europe into one common schema, and also to providesemantic integration along different axes (e.g. subject, space, time). This integration intends toprovideusefulanduser-friendly informationservices forarchaeology.Theservicesare intendedtobeavailablenotonlytoresearchersandrelatedstakeholders,butalsotoawiderrangeofpotentialusersrequiringaccesstocollectionsanddatasets.
It was this user-centred approach that was adopted for the testing of the infrastructure and isreported here. The testing comprised a survey of content-providing partners and formal testingmethods.EachstrandaimedtobalancetherequirementssetoutinD12.1withtheservicesdesignedandimplementedinD12.2andD12.3.FollowingonfromtherecommendationsandresultsofD12.4(which are addressed below) this document tests the infrastructure using broadermethodologiesincludingtheapplicationofusabilityheuristicsandtheSystemUsabilityScale(SUS).Thisdocumentreports on the results of an online survey of content-providing partners and indicates that whiletherewasagenerallevelofsatisfactionwiththeinfrastructurethereareanumberofareasinwhichimprovements would likely benefit the opening out of the infrastructure to an even broaderaudience,particularlyinrelationtouser-friendlinessandhelpdocumentation.
Results of D12.6 testing are presented according to the methods used with further subdivisionsaccording to the services tested. Comments and further recommendations are provided in eachsection.
It must be noted that some of the comments provided in the survey do not concern theimplementation level requiredby theprojectGrantAgreementor the requirements setout in theinfrastructureinitialdesign.Suchcommentsratherconcernfeaturesthatcouldimprovethesystemusability andeffectiveness, andwill be taken intoaccount in thenext version.Asdiscussed in theconclusions,thereisspaceforimprovementinfutureversions,notunexpectedlyforacomplexandnovelsystemastheARIADNEone.
ARIADNED12.6FinalTestingReport
5
1 IntroductionandObjectives
This document is a deliverable (D12.6 Final Testing Report) of the ARIADNE project(“Advanced Research Infrastructure for Archaeological Dataset Networking in Europe”),whichisfundedundertheEuropeanCommunity'sSeventhFrameworkProgramme.D12.6isassociatedwithTask12.4TestingwithinWP12,whichistitledImplementingInteroperability,andfallswithinthelargerARIADNEconceptualframeworkfortheARIADNEe-Infrastructure.ItisanextensionofD12.4Initialinfrastructuretestingreport,andreportstheoutcomesofthe final acceptance testing, taking on board the outcomes of the initial testing. ThisdeliverablewasalsousedtoinformTasks12.3Implementingintegrationand12.4Testing,tofacilitatefinetuningoftheinfrastructurebeforecompletionoftheproject.
Theobjectivesofworkpackage12are:
• ToadaptinfrastructuresprovidedtoARIADNEforintegration.• Todesignandsetup thenecessary tools (crosswalks,mappings)and resources for
interoperability.• Tosetuptheinternal(APIs)andexternal(human)interfacestoaccesstheintegrated
resource.
WorkPackage12iscomprisedofthefollowingtasks:
• 12.1Userequirements• 12.2DesignandSpecifications• 12.3ImplementingIntegration• 12.4Testing
Task12.4testedtheintegratedinfrastructurebuilt inTask12.3 ImplementingIntegration,based on the requirements and design specified in Task 12.1Use Requirements and12.2InfrastructureDesign.Infrastructuralelementswithuserinterfacesweretestedforthistask.
Documented in D12.2, the ARIADNE Registry is a key element of the interoperabilityframework and drives resource discovery within the public-facing ARIADNE portal. Inparticular, the deposit service allows registered users to deposit data and metadatadescriptionsfollowingtheARIADNEDatasetCatalogueModel(ACDM)schema.TheprovidedmetadataismanagedthroughaDigitalAssetsManagementservice,andispresentedtothepublic through theARIADNEportal. TheResourceDiscoveryServices (mainly indexingandretrieval) enables access to data resources and integrated viewing of data resourcedescriptions, through the portal. The vocabulary management service is responsible formaintaining a list of SKOSified vocabularies and thesauri. The metadata enhancement
ARIADNED12.6FinalTestingReport
6
service allows for automatic enhancement of metadata found in ACDM records. Theseenhancements include the mining of relations, and automatic linking with thesauri andvocabulariesetc.[D12.2,10]
Insummarythearchitectureprovides:
1.thefunctionalitiesoftheARIADNECatalogue;
2.interoperabilityandintegrationfunctionalitiesforarchaeologicalresources;
3.theplatformuponwhichtheARIADNEserviceswillrun.[D12.2,11]
Thisreport is focusedontestingeachcomponentoftheARIADNEinfrastructuretoensurethateachmoduleiscompatibleandcontributestotheintegratedsystem.Themethodsusedto test the ARIADNE integrated infrastructure were varied and included group meetingsrelating to portal testing, communication by e-mail/skype with individual partners aboutimplementing the recommendations of D12.4, an online survey of content-providingpartnersaswellasin-housetestingusingformalusabilitymethods.Thisreportalsooutlinesthe current workflows in place for providing content for integration into the ARIADNEinfrastructure.
ARIADNED12.6FinalTestingReport
7
2 UseRequirements,ServiceDesignandImplementation
TheworkofWP12 specifieduse requirements for theARIADNE infrastructure (noting theextensiveworkofWP2onuserneedsandWP3ontheRegistry),todesignandimplementaspecification for the infrastructure based on the use requirements. Use requirementsspecified in D12.1 focused on three areas: datasets, metadata standards, schemas andvocabularies and access and sharing policies. The use requirements reflected the dataavailableinthearchaeologicalsector,andprioritiesfortheintegrationactivitiesfordifferenttypes of data. Use requirements for the infrastructure have both practical and policyimplications,bothofwhichmustbereflectedintheimplementationasappropriate(D12.1,4-5).Thedesignspecificationfocusedonanintegrationstrategywhichaddressedmetadataintegrationanddataintegration,andinthefirstinstance,enabledcross-searchofresourcesthroughWhat,Where,When and Resource Type facets. Consistency and quality are keyrequirementsforcontentaddedtotheARIADNEinfrastructure(D12.2,16-17).
D12.4providedacomparisonoftheuserequirementsofD12.1,theservicedesignofD12.2anditsimplementationinD12.3.Italsoofferedpreliminarytestingofthevariousinterfacesof the infrastructure which were available at the time. This document provides furthertestingof the infrastructurevia severalmethodswhich incorporateaspectsof thegeneraluserequirementsofD12.1withadditionaltestingusingformalmethods.
ARIADNED12.6FinalTestingReport
8
3 InfrastructureDesign
3.1 InfrastructureElements
TheintegratedinfrastructureincorporatestheARIADNEregistry,aggregationandvalidationservices (MORe) as well as the ARIADNE portal. These combined elements enabledarchaeological and cultural heritage content in various forms to be made available toresearchers,chieflythroughtheirintegrationintheonlinewebportal.Althoughtheportaliscrucialtodiscovering,visualisingandaccessingarchaeologicaldata,evenmorecrucialisthequalityof themetadatasubmittedto the infrastructurebyARIADNEpartners.This isbornoutincommunityfeedbackandconcernsarticulatedinD2.1andacrossotherdisciplinesandinfrastructures (see for example metadata principles in the Framework of guidance forbuilding good digital collections of the National Information Standard Organisation). D2.1reported that 50% of archaeological researchers felt that metadata quality was ‘veryimportant’whileafurther34%felt itwas ‘rather important’. The infrastructuresoughttomitigatemetadata quality issues by validating deposited data against the ACDM schema,thus ensuring aminimum standard which could be used for integration. The enrichmentservices aim to further improvemetadata quality by allowing automatic enhancement ofACDM records. To continue to ensuremetadata quality the workflow for how the datareachestheportalmustbeclearandwell-definedforfuturecontentproviders.
ARIADNED12.6FinalTestingReport
9
Figure1.TheARIADNEdataaggregationworkflowfromD12.3[p.13]
3.2 ContentProvisionWorkflow
ARIADNE content-providing partners have different routes for sharing their data in theinfrastructure,theseareoutlinedbelow.
Stage1:Ingest
Themetadataregistryprovidesthreedifferentinterfacesformanaginginformation:
1. WebbasedUI.This isdesigned formanualeditingof informationbyhumanusers.Eachcontentproviderhastheirownuseraccountthroughwhichtheycanaddnew/edit existing informationper class and linkdifferent instances together. TheuseofthistoolwasextensivelydocumentedinD12.4,seepp.18-33.
2. Batch Import. The batch import is available to partners through contact withtechnicalpartnersandallowsthebatchingestofdatasetsandcollectionsthrough:a)OAI-PMH target, b) XML file, c) Excel file. Templates for Excel and XML files areavailable through the Registry support portal andmust be correctly structured foringest.TheExcelfilecontainsallmandatoryclasses,aswellassomerecommendedclasses,andisthensenttotechnicalpartnersforinclusionintheregistry.Similarly,providerscancreateanXMLfilefortheirdata,thiscanbevalidatedagainstan.xsdfilealsoavailableintheregistrywebtoolandthensenttotechnicalpartners.
ARIADNED12.6FinalTestingReport
10
3. REST API. The REST API allows users to export all catalogue information (perprovider)inXMLformatandtomanipulateitthroughaRESTAPI.TheaccesstotheAPI is controlledbya sessionkey that is generated fromwithin theWeb interface(perprovider).RESTAPIcanbeused to return,edit,addanddelete records in theARIADNECatalogue.Theauthenticationrequiredtoperformthesetasksisprovidedthrough the ARIADNE Catalogue, using API keys for the machine interface anduser/passwordchallengeonthehumaninterface(D12.2,34).
Stage2:Validation,enrichmentandpublication
TheMOReaggregationworkflowhasbeenproventhroughexperienceinLoCloud,CARAREand other European projects.MORe provides a flexible architecture, and is used here toaggregatecontent invarious formats,andfromavarietyofsources, totheRDFstore, theregistry and Elasticsearch. MORe incorporates a micro-service oriented architecturesupportingthefollowingcoreservices(D12.3,13):
• Input• Validation• Transformation• Enrichment• Publication
The toolhasauser-friendly interface formetadata ingest, validationandenrichment.Theworkflowforthetooliswelldesigned,astheuserisguidedthroughthestepsnecessaryforcompleting the publication ofmetadata to the portal by a virtual assistant. An illustratedworkflowisgiveninAnnexII.
After validation and enrichment, data is then published to theweb portal for search anddiscoveryaswellasintegratedviewingofdataresourcedescriptions.
ARIADNED12.6FinalTestingReport
11
4 Usability&Testing
4.1 ResultsofD12.4
TheinitialtestingoftheinfrastructurewasreportedinD12.4,theresultofwhichwasalistof specific recommendationswhichwere circulated to relevant partners.Upon consultingwithtechnicalpartnersitwasagreedthatthewebtoolcomponentoftheregistrywouldnotbedeveloped furtherowing to its limitedusebycontentproviders. Itwas recommended,however, that should the tool be used more frequently in the future by new contentproviders, the recommendationsofD12.4 relating to theserviceshouldbeaddressedandincorporatedintothewebinterface.Therecommendationsrelatedtootheraspectsoftheinfrastructurearelistedbelow,withtheadditionofacommentsfieldindicatingthecurrentstatus of the recommendation. These comments were informed by recent testing of theinfrastructureaswellaspartnerfeedbackfromthesurveyconductedforthisdeliverable.Itmustbeobservedthatmanyoftheissuesevidencedconcernfuturenewusers,andthuswillneedtobeconsideredwhenplanninganewreleaseofthesystem,wherecontentprovisionwouldbeenlargedtoamuchwideraudience.
Recommendations Comments
3 12.1Requirements
R31 3.1.1
Resolvetheuseoflanguagearoundtheterms‘Registry’and‘Catalogue’inanysupportdocumentationanduserinterfaces.
To be addressed in future versions fornewcontentproviders.Currentusersareclear about the equivalence of the twoterms.
R32 3.1.1
ClarificationoftheuserequirementsforthePreviewService.
Clarified:Previewservicewillnotcontainapreviewofdatabutwillprovideapreviewofmetadataintheformofsearchresultsintheportal.
ARIADNED12.6FinalTestingReport
12
Recommendations Comments
R33 3.1.2
Appropriate,targetedguidancefortheuserintheuseoftheRegistrytoadddataresourcesisrequired.
Tobeaddressedinfutureversions.Althoughthereisuserdocumentationavailable,thereshouldbeonesinglesectionontheportalinterfacewhichallowsuserstodeterminehowtheymightdepositdata,includingbasicworkflow,ingestformatsandcontactdetailsforpartnersresponsibleforthis.
R34 3.1.2 Abusinessandsustainabilitymodelisrequired.
Ongoing:Thisisbeingexaminedaspartofthesustainabilityplan.
R35 3.1.3 Guidanceforusersonthenatureoftheerrorsencounteredduringthedepositandingestprocesswouldbehelpful.
Partiallyresolved,andtobecompletelyaddressedinfutureversions.TherearenofurtherdevelopmentsconceivedfortheRegistrytool.Errorsmanifestedthroughotherdeposit/ingestprocesses(e.g.Excel,XML)arecommunicatedtoprovidersdirectlybytechnicalpartners.TheMORetoolhasrobusterrornotifications.
R36 3.1.3 Awordcloudofthe‘derivedsubject’facetorGettyAATterms,eitherasafilterorasecondwordcloudinthe‘What’browsingarea.
Resolved:ThishasbeenresolvedbyWP13andthewordcloudisnowpopulatedwithGettyAATterms.However,thisleadstoweightinginfavourofthosedatasetswhichhavebeensuccessfullyenrichedbyGettyAATterms,andisperhapsnotfullyrepresentativeoftheentiretyofthedatainthecatalogue.
R37 3.1.4 Theprojectshouldcommunicateclearlytousersitsapproachtodataandmetadata.
Resolved:Whilesomeuserswereuncertainastowhetherdatawastobeingestedatitemorcollectionlevel,partnersareclearthatthecataloguecontainsmetadatawithexternallinkstodata.Experimentsfordataintegrationatitemlevelareongoingandaretakingplacewithinalternativemetadatarepositories(CNR-ISTI).Thesewillbeimplementedinaseparateexperimentalsectionoftheportalinterface.
ARIADNED12.6FinalTestingReport
13
Recommendations Comments
R38 3.1.5 ThesourcevocabulariesavailablethroughtheVocabularymatchingtoolcouldbeexpandedtoallowbroadermappingtoAAT.
Resolved:CurrentlythevocabularymatchingtoolcontainsvocabulariesfromBritishproviders.Asthetoolisofferedasapartnerservice(asopposedtoonedevelopedspecificallywithinARIADNE)itisnotpracticabletoexpandthesourcevocabularies,inadditionmostpartnershavemappednativevocabulariestotheGettyAATaspartofWP15.
R39 3.1.5 Guidancedocumentationonhowtopreparevocabulariesforinclusioninthetoolandformappingwouldberequired.
Resolved:PartnerscommunicateddirectlywithWP15partnersandmappedvocabulariesareavailablethroughMOReforenrichmentofproviderdata.
R40 3.2.1 DocumentationisrequiredforthePeriodOenrichmentservice.
Resolved:ThePeriodOenrichmentisinbeingimplementedinMORe.
R41 3.2.5 Clarify,fromD12.1andD12.2,whatkindofintegrationandinteroperabilityissoughtforscientificdatasets.
Resolved:ThisaspectisbeingaddressedbyWP16wheredataminingisongoingfordendrochronologicaldata.
5 Aggregationservices
R42 DocumentationfortheuseofMOReshouldbeavailablewithintheSupportPortalandaccessiblefromwithinMORe.
Postponed.SomepartnersexpressedconfusionabouttheroleofMOReintheworkflow.InfutureversionstherewillneedtobeclearguidelinesonpartneraccesstoMORe.
R43 IncludeworkedexamplesofprovisionoftheCollectionclassinuserguidance,includingmetadataprovisionfortheitemsordataresourcesofwhichthecollectioniscomposed,wouldillustrateforusershowtheircontentwillberepresentedintheportal.
Postponed.Severalsurveyrespondentsproposedthatfurtherexamplesbeprovidedintheuserguidanceinfutureversions.
ARIADNED12.6FinalTestingReport
14
Recommendations Comments
R44 FurthertestingshouldbecarriedoutwithapanelofuserswhodonothavepreviousexperienceofMOReandassociatedtools.
Resolved:Seeinternaltestingsectionofthisdocument.
R45 ManualentryworkflowintheRegistryshouldbeharmonisedwithuserguidanceandthelatestversionoftheACDM.
Resolved:Nofurtherdevelopmentofthetoolinthenearfuture.Thismaychangeifitisusedmorefrequently.Awalk-throughvideoofthetoolmaybeusefulforfuturefirsttimeusers.
6 SupportPortal
R46 Userguidanceshouldincludeexamplesofalldataresourceclassesworkedwiththeirlanguageresourceassociations.
Postponed to future versions. See D12.4pp38-9.
R47 ExpandtheRegistryFAQ. Postponedtofutureversions.QuestionsintheRegistryFAQarequitespecific,moregeneralquestionsshouldbeadded,pageneedstobere-structuredtoincludesectionsforease-of-use.
R48 Changenavigationaroundaddinganinstanceoffoaf:agentfromwithinadataresource.(Seerecommendationunder‘AddDataset’4.1.3above).
Postponedtofutureversions.Eventhoughthistoolhasceasedactivedevelopment,thisfeatureshouldbeaddressedinthefutureasapotentiallysignificantusabilityissuefornewusers.
R49 Comprehensivestep-bystepworkflowandmappingguidancecouldbeprovidedbothinoneresourceandbrokenintopartswhichreflectvarioususerinteractionswiththeinfrastructure.
Postponed to future versions. This wasremarked on by only one respondent ofthe online survey. They noted that thesupport portal needs ‘to be edited forusers not familiar with the ARIADNEproject, and with their ease of use inmind’,suchaspotentialnewcontributorsthe ARIADNE registry envisaged in thefuture.
Table1ListofRecommendationsfromD12.4Initialinfrastructuretestingreport.
ARIADNED12.6FinalTestingReport
15
4.2 TestingMethodologies
There are several methodologies for testing infrastructural systems like the ARIADNEregistry and portal. D12.4 outlined further methods for testing, including theimplementationoftheSystemUsabilityScale(SUS)inthetestingoftheregistry(D12.4,40).Thisisoneofthemostcommonmethodsforusabilityandisdiscussedindetailbelow.
Other usability methods include inspection and test methods (see table 2). Inspectionmethodsarecarriedoutbysystemcreatorsanddonotuseendusersaspartofthetestingprocess. The methods included under this type are heuristic evaluation, cognitivewalkthroughandactionanalysis.Thesecondtype;knownastestmethods,involvetheuseofendusersinthetestingphaseandalsoencompassesthreestrands,thinkingaloud,fieldobservationandquestionnaires(Holzinger,2005).
Inspection methods seek to improve interface usability by checking design againstestablished standards (Holzinger, 72). Studies have shown that inspectionmethods oftenfindmany issues that are overlookedby testingwith real users, and likewise user testingillustratedproblemsthatweremissedininspectionmethods,thereforethemostbeneficialcourseistotestusingacombinationofinspectionandusertestmethods(Nielsen(1994a),413).
ARIADNED12.6FinalTestingReport
16
Inspectionmethods Testmethods
Heuristicevaluation ThinkingAloud
CognitiveWalkthrough FieldObservation
ActionAnalysis Questionnaires
Table2UsabilityEvaluationTechniques
It is commonly agreed that there are five essential usability characteristics: learnability,efficiency, memorability, low error rate and satisfaction. The performance of a systemrelative to these characteristics ensures the systemwill be successful to endusers. JakobNielsenextendedusabilityprinciplesandoutlinedten“heuristics”orprincipleswhichuserinterfacesshouldcomplywith,seebelow(Nielsen,1994b,25-62).
For the purposes of testing the ARIADNE integrated infrastructure, a combination ofmethodswasused.Fromtheinspectionmethodstier,heuristicevaluationwasusedasthemosttime-efficientmethod,whilefromthetestmethodstier,aquestionnaireintheformofanonlinesurveywasconducted.
Heuristicevaluation
“Heuristicevaluationisausabilityengineeringmethodforfindingtheusabilityproblemsinauserinterfacedesignsothattheycanbeattendedtoaspartofaniterativedesignprocess”(Nielsen,1994b,25-6).
Inthis inspectionmethod,ausabilitytestercarriesoutseveralrun-throughsofthesysteminterfaceand tests themagainsta setofdefinedprinciples.Themostcommonlyusedsetare the ‘heuristics’ of Jakob Nielsen which set out ten standards against which systemsshould be tested.Nielsen’s heuristicswere used for testing the interface of theARIADNEintegratedinfrastructure.Heuristicevaluationisacostandtimeeffectivemethodfortestingand has been noted to have a low “intimidation barrier” to its implementation (Nielsen,1994b,25).
ARIADNED12.6FinalTestingReport
17
4.3 InternalTesting
4.3.1 Methodology
Each individual evaluator inspects the interface alone. After evaluations have beencompleted,findingscanbeaggregated,butpriortothis,thereisnocollaborationbetweenevaluators inordertoensure independentconclusions(Nielsen,1994b26).Astheregistrywasdeemedtohavetoosteepalearningcurveinthistestingphase,onlyusersfamiliarwiththetoolassessedthisaspect.Abetterindicationofuserexperiencewiththeregistrycanbegaugedfromthesurveyquestionnairecirculatedtocontent-providingpartners(seebelow).
4.3.2 Testing
Agroupofinternaltesterstestedtheinfrastructureindividually,astheywerefamiliarwiththe overall aims of the ARIADNE project and experienced in both content provision andutilisation of the tools. Each user was given a form outlining the usability principles andaskedtogivethemarating(•Poor/••Moderate/•••Good).Thetesterswerealsoaskedtoprovide comments on aspects of the infrastructure that addressed, or didn’t address, theusabilityprincipleinquestion.
Feedbackfromtesterswasamalgamatedandrelevantcommentsprovidedbelow.Asusual,theyconcernimprovementstobeimplentedinfutureversions.Table3showstheusabilityprinciplewithcorrespondingdescription.
Visibilityofsystemstatus
Thesystemshouldalwayskeepusersinformedaboutwhatisgoingon,throughappropriatefeedbackwithinreasonabletime
Matchbetweensystemandtherealworld
Thesystemshouldspeaktheuser’slanguage,withwords,phrases,andconceptsfamiliartotheuser,ratherthansystem-orientedterms.Followreal-worldconventions,makinginformationappearinanaturalandlogicalorder.
Usercontrolandfreedom
Usersoftenchoosesystemfunctionsbymistakeandwillneedaclearlymarked“emergencyexit”toleavetheunwantedstatewithouthavingtogothroughanextendeddialogue.Supportundoandredo.
Consistencyandstandards
Usersshouldnothavetowonderwhetherdifferentwords,situations,oractionsmeanthesamething.Followplatformconventions.
ARIADNED12.6FinalTestingReport
18
Errorprevention Evenbetterthangooderrormessagesisacarefuldesignwhichpreventsaproblemfromoccurringinthefirstplace.
Recognitionratherthanrecall
Makeobjects,actions,andoptionsvisible.Theusershouldnothavetorememberinformationfromonepartofthedialoguetoanother.Instructionsforuseofthesystemshouldbevisibleoreasilyretrievablewheneverappropriate.
Flexibilityandefficiencyofuse
Accelerators-unseenbythenoviceuser-mayoftenspeeduptheinteractionfortheexpertusertosuchanextentthatthesystemcancatertobothinexperiencedandexperiencedusers.Allowuserstotailorfrequentactions.
Aestheticandminimalistdesign
Dialoguesshouldnotcontaininformationwhichisirrelevantorrarelyneeded.Everyextraunitofinformationinadialoguecompeteswiththerelevantunitsofinformationanddiminishestheirrelativevisibility.
Helpusersrecognize,diagnose,andrecoverfromerrors
Errormessagesshouldbeexpressedinplainlanguage(nocodes),preciselyindicatetheproblem,andconstructivelysuggestasolution.
Helpanddocumentation Eventhoughitisbetterifthesystemcanbeusedwithoutdocumentation,itmaybenecessarytoprovidehelpanddocumentation.Anysuchinformationshouldbeeasytosearch,focusedontheuser’stask,listconcretestepstobecarriedout,andnotbetoolarge.
Table3UsabilityheuristicsasoutlinedbyNielsen.[Source:J.Nielsen,
'HeuristicEvaluation'Usabilityinspectionmethods17.1(1994)p.30.
ARIADNERegistryTool
Ratings:•Needingimprovement/••Satisfactory/•••Good
Usabilityfeature Rating Comments
Visibilityofsystemstatus • Itwassometimesunclearwhendatahadbeensuccessfullyingestedintotheregistry.Itisrecommendedthatinfutureversionsusersaregivenfeedbackonsuccessfulorunsuccessfulactionsinthetool.
Itisdifficulttodeterminewhereintheprocessoneis,whetherthesystemissavingprogressandthestatusfordatainputatorganisationlevel(multipleseparateaccountsforindividualsinoneorganisation,withoutflagsofothercontentorvisibilityacrossaccounts.)
ARIADNED12.6FinalTestingReport
19
Matchbetweensystemandtherealworld
•• Somehelptextscouldbeimprovedtoallowinformation‘appearinanaturalandlogicalorder’.
Usercontrolandfreedom •• Thereareadequateexitpointsforusers,howeverthereisscopetosupportundoandredo.
Consistencyandstandards •• ForthemostpartthetoolfollowedthewordingandstructureoftheACDMwhichledtoastandardmethodofexpression,however,itisnecessarytopointuserstotheACDMspecificationshouldtheyneedtochecktheirmetadataagainstthemodel.ThiscouldbeintheformofalinktothedescriptionoftheACDMclass,propertyorattributeinthespecification.
Errorprevention •• Therewerefewexperiencesofdirecterrorsinthetool,forspecificexamplesseeD12.4,p.22.
Requiringvaluesincertainfieldsreducesthelikelihoodoferrorsoccurringinlaterstagesofthedataingestionprocess.
Recognitionratherthanrecall •• Optionsandactionsarevisibletotheuserinthetoolwiththesameinterfaceforeachresourcetobeaddedwhichfacilitatesuserrecognition.
Flexibilityandefficiencyofuse • Bothexpertandnoviceusersinteractinthesameway,thereisonlyoneinterfaceforallusers.Thereisnopossibilitytotailorfrequentactions.
Aestheticandminimalistdesign •• Althoughtheuserinterfaceisminimalist,itmaybenefitfromaclearervisualstructure.Sevenoftheavailableresourcetypesarenotvisibleonthispage,includingthemandatoryfoaf:agent.Keydataresourceclassescouldbegroupedtogether,andlanguageresourcesgroupedseparately,reflectingtheresourcetypes.Abasicvisualisationmayhelptheusertobetterunderstandtherelationshipsbetweendataresourcesandlanguageresources,andillustratewhichlanguageresourcesaremandatoryorrecommendedforparticulardataresources.
Helpusersrecognize,diagnose,andrecoverfromerrors
•• Therewerenoexperiencesofcodederrormessages,however,thelackoffeedbackoncedatasetshavebeenaddedandsavedcreatesuncertaintyaboutwhetherresourceshavebeensuccessfullyadded.Apop-upmessageindicatedthatresourcehasbeensuccessfullysaved/ingestedwouldbeuseful.
Helpanddocumentation •• ThereisahelpsectionontheRegistrybutthiscouldbeexpanded.Itisimportanttoincludeawayofcontactingpartnersfortechnicalsupport.However,whentechnicalsupportwasrequiredresponsewasswiftandefficient.
ARIADNED12.6FinalTestingReport
20
AggregationService,ValidationServices,MetadataEnrichmentServices:
theMOReaggregationtool
Ratings:•Needingimprovement/••Satisfactory/•••Good
Usabilityfeature Rating Comments
Visibilityofsystemstatus ••• Thetoolhasverygoodfeedbackmechanisms.Usersreceivealertseachtimetherehasbeenachangeinthestatusofapackage(ingestion,validation,enrichmentetc.)
Matchbetweensystemandtherealworld
••• Thetoolissimpletouse.Avirtualguidetakesusersthroughthesystemwithsimplepop-upinstructionsonceeachstagehasbeencompleted.Languageandconceptsareplainandeasilyunderstandabletonon-technicalusers.
Usercontrolandfreedom ••• Supportsundo:Packagescanbeeasilyrejectedordeletedatanystagepriortoandafterpublication.Thereareclearexitpathsforusersatallstageswhenusingthetool.
Consistencyandstandards •• Languageusedinthetoolisuniformthroughout.However,thereissomeneedtoupdatethesupportdocumentationinthemicro-servicessectionwhichcurrentlypointstoLoCloudsupport.
Errorprevention ••• Therewasnoexperienceofdirecterrorsinthetool.
Recognitionratherthanrecall ••• Thetoolhasaveryclearvisualstructurewitheachcomponentaccessiblefromthemaininterface.Objects,actionsandoptionsarevisibleatalltimesandremainsofromonepartofthedialoguetoanother.
Flexibilityandefficiencyofuse ••• Frequentactionscanbetailoredbytheabilitytobuildanenrichmentplanfordata,thiscanbere-usedfornewlyingesteddata.
Aestheticandminimalistdesign ••• Thetoolhasaveryclearinterfacedesign.
Helpusersrecognize,diagnose,andrecoverfromerrors
••• Messagesareeasytounderstandfornon-technicalusers.Validationerrorsareclearlyindicated,withusersabletogetdetailsfortherecordsandelementsthatarecausingtheerrors.Anotificationlogfrompreviousactionscanalsobeaccessedinthetool.
Helpanddocumentation •• WhilethereissupportdocumentationavailablewithintheMORetool,thisisfortheLoCloudproject.ThisdocumentationshouldtailoredforandbemadeavailablefortheARIADNEinstanceofthetool,oratleastbedescribedontheARIADNEwebsite.
ARIADNED12.6FinalTestingReport
21
Portal1
Usabilityfeature Rating Comments
Visibilityofsystemstatus ••• Itiseasytoseewhereyouareinasearch,asthefacetschosenwhichleadtocurrentviewsaredisplayedonscreenbesidethesearchresults.
Matchbetweensystemandtherealworld
••• Therearenosystem-orientatedtermsandtheportalfollowsreal-worldconventions.Itmaybenecessarytore-structureandre-writethedescriptionsonthesupportpagefornon-technicalusers.ForexampleitisnotimmediatelyclearwhytheACDMisdescribedhere.Itwouldbemoreusefultohaveaguidetotheportal,ratherthaninformationabouttheregistry.
Archaeologists’severalpathstodataandinformationarewellreflectedintheportal.Extensiveeffortsinintegratingarchaeologicalvocabulariesandexistinginformationstructuresensureagoodmatchbetweenrealworldarchaeologyandthesystem.
Usercontrolandfreedom ••• Itiseasyforuserstorefinetheirsearchparametersusingthefilters,thesefacetscanbeaddedorremovedwithease,thussupportingundoandredo.
Alinktothesearchpagetobeginagainisalwaysavailable.
Consistencyandstandards •• Informationdisplayisclearandconsistent.WhileARIADNEhasastrongdatastandard,ensuringadherencetothestandardacrosspartnersfromdifferentcontextsiskeytoconsistencyacrossarchaeologicaldataintheportal.
Theremaybesomeconfusionamongusersinthenumberoffacetsrelatedtothematicconcepts-inparticularthepresenceofthreesimilarrefinements-subject,keywordandoriginalsubject.Itshouldbemadeclearthat
1Notetheportalhasalsobeentestedtogetherwiththeservices inWP13,readersareurgedtoalsoconsultWP13deliverables.
ARIADNED12.6FinalTestingReport
22
‘subject’relatestotheGettyAATterm,whiletheotherfacets(fromthenativemetadata)maybenefitfromahelptexttopointuserstowherethesetermsoriginate(i.e.withdepositors).
Errorprevention ••• Theportalalwaysreturnssearchresultsratherthanproducingerrormessages.Itisalsoclearwhentherearenoresultstodisplayforagivensearch.
Recognitionratherthanrecall ••• Theresourcelandingpagesandsearchpagesdisplaybuttonsforrelevantoptionsinauniformway.However,nogeneral‘help’buttonisvisiblewithintheportal.
Flexibilityandefficiencyofuse •• Useoffacetedsearchingmayincreaseefficiencyofuseforexperiencedusers.Optionstosaveshortcutsforfrequentactionsarenotavailable.
Itwouldbeworthexploringtheadditionofaloginsothatuserscancreateanaccountintheportalandsaveorexporttheirsearchresults.
Aestheticandminimalistdesign ••• Theportalverysuccessfullysharessufficientrelevantinformationandoptionswithoutcrowdingthesearchorlandingpages.
Someoftherollingimagesonthelandingpageoftheportalmayneedtobereviewedasthereneedstobeahighcontrastbetweenthesearchandbrowseboxesandthebackgroundimages.Someimagesaretoo‘busy’andneedtobeahigherresolution,filesizesshouldbereviewedasoccasionallycertainimagestakelongertoloadthanothers.
Help users recognize, diagnose, andrecoverfromerrors
•• Therewereveryfewexperiencesoferrorsintheportal.Thefunctionforviewingarecord’smetadatainACDMXMLisnotcurrentlyfunctioning.
Helpanddocumentation •• HelpanddocumentationisnoteasilyvisibleoraccessiblewithinthePortal.TheARIADNEPortalGuidePDFdocumentisaclearandeffectiveguidewhichshouldbeaccessiblefromtheportal‘about’section,ratherthantheprojectwebsite.
ARIADNED12.6FinalTestingReport
23
4.4 OnlineSurvey
Asurveywasconductedamongcontent-providingpartnerstotheARIADNE infrastructure.Thiswasdonethroughanonlinesurveytool (SurveyMonkey)andthe linkcirculatedtoallpartners through basecamp. There was a good response rate from content-providingpartners,withabout65%ofpartnerscompletingthesurvey.ThequestionnaireimplementsausabilitytechniquefromthesecondtierofTable2,balancingtheheuristicevaluationusedabove.
Themain impetusbehindthesurveywastotestthe integratedfeaturesandworkflowsoftheARIADNEinfrastructure.Assuchitposedquestionsontheregistry,enrichmentservicesandtheportal,withafocusoncontentpreparation,ingestandpublication.Therewasalsosignificant examination of the help/support documentation available to users. HowintegratedtheoverallarchitectureiswillbeessentialforthefutureintegrityoftheARIADNEproject.Thesurveyprovidedmultipleopportunitiesforpartnerstoexpandontheiranswersor comment on aspects of the infrastructurewhich they felt were not highlighted in thesurveyitself.
The survey results arepresentedherebasedon the various sectionsof the infrastructurewhich were evaluated. The results of the survey were communicated to the technicalpartnersforreviewandanalysis,andforeventualintegrationintotheinfrastructure.Thisisanongoingprocessasmorecontentisingestedandtheinfrastructureisrefined.
A copy of the survey questions is available in Annex I. Within the survey there was anembeddedusabilitytestingmechanismcalledtheSystemUsabilityScale(SUS).
4.4.1 SystemUsabilityScale(SUS)
The SystemUsability Scale (SUS)was created by John Brooke in 1986 and is designed toprovidethefollowingbenefits:
• Isaveryeasyscaletoadministertoparticipants• Itcanbeusedonsmallsamplesizeswithreliableresults• Isvalid-itcaneffectivelydifferentiatebetweenusableandunusablesystems
TheSUSmustbeunderstoodintermsofthelimitationsofusabilityasatangiblequalityandmustbeviewed‘intermsofthecontextinwhichitisused,andtheappropriatenesstothatcontext.’Itis,asBrookeoutlined,‘impossibletospecifytheusabilityofasystem…withoutfirstdefiningwhoaretheintendedusersofthesystem,thetasksthoseuserswillperformwith it, and the characteristics of the physical, organisational and social environment in
ARIADNED12.6FinalTestingReport
24
which itwill be used.’2 (Brooke, 1996) It has been found to be highly reliable and usefulacrossarangeofinterfacetypes(Bangor,KortumandMiller,2008).
TheSUSissimpleten-pointscalewhichaimstogiveaglobalviewofsubjectiveassessmentsof usability. Itwas born of the need to replace hitherto complicated evaluationmethodswhichwere time-consuming, expensiveand frustrating forusers. The ten statements thatcomprisetheSUSaredesignedtocoveravarietyofdifferentaspectsincludingtheneedforsupport, training and complexity and therefore ‘have a high level of face validity formeasuringusabilityofasystem.’
SUSstatements:1. IthinkthatIwouldliketousethissystemfrequently2. Ifoundthesystemunnecessarilycomplex3. Ithoughtthesystemwaseasytosue4. IthinkthatIwouldneedthesupportofthetechnicalpersontobeabletousethis
system5. Ifoundthevariousfunctionsinthissystemwerewellintegrated6. Ithoughttherewastoomuchinconsistencyinthissystem7. Iwouldimaginethatmostpeoplewouldlearntousethissystemveryquickly8. Ifoundthesystemverycumbersometouse9. Ifeltveryconfidentusingthesystem10. IneededtolearnalotofthingsbeforeIcouldgetgoingwiththissystem
Respondentsareaskedtorateeachof theabovestatementsona fivepointscalerangingfromstronglydisagree(1)tostronglyagree(5).
SUSScoring
SUS scores are calculated by totally the score contributions from each statement, eachstatement’s score ranges from0-4. For statements1,3,5,7and9 the score is the ratinggivenbytherespondent-1.Forstatements2,4,6,8and10thescoreis5–theratinggivenby the respondent. The total of all the scoresmultiplied by 2.5 gives the overall systemusabilityvalue.Thisrangesfrom0to100.
TheaverageSUS score is 68, scoresabove this are considered ‘aboveaverage’. The scoreranges from0 (negative) to100(positive).However, theSUSscore isnotapercentage. InadditiontothescoringBangor,KortumandMillerhavedevelopedanadjectiveratingscaletoprovidemoremeaningfulandhelpfulresults.Theseareshowninthechartbelow.
2TheseaspectshavebeenaddressedinthecourseoftheARIADNEprojectbyseveralstrands,inparticularbytheworkofWP2andtheresultingD2.1FirstReportonuser’sneeds.
ARIADNED12.6FinalTestingReport
25
Figure2.MeanSUSscoreratingscorrespondingtothesevenadjectiveratings(errorbars+/-onestandard
errorofthemean).FromBangor,KortumandMiller,‘DeterminingwhatindividualSUSscoresmean:addingan
adjectiveratingscale’inJournalofUsabilityStudies,4:3(2009)
SUStestingwascarriedoutthroughtheTask12.4surveyfortheregistryandportal;resultsaregivenbelow.
Registry
Results for the registry varied somewhat frompartner to partner (see comments below).The lowest SUS scorewas 37.5with thehighest being75.On average the SUS scorewas58.75,belowwhatisconsideredaverage.Itgivesanadjectiveratingof‘Ok’.
Portal
TheportalhadconsiderablyhigherscoringthantheRegistry.TheSUSscoresfortheportalranged from 42.5 (from one partner, the next lowest score being 65) to 90 indicating ahigher levelof satisfactionamongpartners.TheaverageSUSscorebeing77.5,wellabovetheaverage.Thisgivesanadjectiveratingof‘good’.
Comments: While the results indicate that there is general satisfaction with how theinfrastructure performs, therewas in each respondent’s case a lower level of satisfactionwith the registry than therewaswith theportal. This isperhapsa falsedichotomyas theregistrycontains thedatapowering thecontent thatappears in theportal. It is likely thatthe lower scoremore fully represents some difficulties experiencedwith adapting nativecontenttotheARIADNEinfrastructurewhichofteninvolvedvariedtechnicalconsiderationsonthebestmethodformapping,ingestingandexposingprovider’sdatabasestotheproject.Duringsystemdevelopment,uploadingcontenttotheregistrywassometimescumbersomeas it happens in earlydevelopment stages,while thesedifficulties substantiallydecreasedwhenthesystemwasfullyoperational.Possibly,thememoryoftheproblemsencountered
ARIADNED12.6FinalTestingReport
26
duringthisinitialdevelopmentstageinconsciouslyinfluencedsomeanswersinthenegative,especiallywithnon-technicalpartners.Theingestionactivitywascarriedoutatapartner-to-partner level and in the initial stages involved an iterative process whereby theinfrastructure was being developed as different (and sometimes unforeseen) types ofcontent was being offered by partners. This (largely undocumented) work is a keyconsideration for future collaborationwithinARIADNEand thepublicationof case studiesforhownativedatarepositorieswereexposedtotheinfrastructuremaybeusefulforfutureusers.
Animportantaspectoftheresponsesisthedifferenceofexperiencefrompartnerswhohadlarge infrastructural resourcesat theirdisposalandthosewhohadsmalleroperationsandfewer technical staff. This is a critical factor to consider when assessing the way theconsortium will operate in the future. Should larger partners share resources (training,workshops, technical services etc.) or should these partners act as conduits for contentprovisionbysmallerproviders?Alternatively,shouldtheinfrastructurebeadaptedtoensurethat smaller institutions with fewer resources are catered for by improving the overallaccessibilitytorelevantdocumentationandtechnicalsupport?
4.4.2 SurveyResults
The five areas defined as important for archaeological research data highlighted by D2.1were:
• Datatransparency:havingagoodoverviewofavailabledatasets• Dataaccessibility:therequireddatasetsareavailableinanuncomplicatedway• Metadataquality:theavailabledatasetsarewelldescribed• Dataquality:theavailabledatasetsarecompleteandwellorganised• Internationaldimension:havingeasyaccesstointernationaldatasets
TheseareaswereassessedinthecourseoftheonlinesurveyconductedforWP12testing.Respondents were asked to rate how they felt different aspects of the infrastructureaddressedthekeyfiveareas.
ARIADNED12.6FinalTestingReport
27
TheRegistry
Figure3.ChartshowingthelevelofsatisfactionwiththeregistryinrelationtothegeneraluserequirementsoftheARIADNEinfrastructure
ThePortal
Figure4.ChartshowingthelevelofsatisfactionwiththeportalinrelationtothegeneraluserequirementsoftheARIADNEinfrastructure
0
10
20
30
40
50
60
70
80
Excellent Good Average Poor VeryPoor
Datatransparency DataAccessibility Metadataquality
DataQuality InternaronalDimension
0102030405060708090
Excellent Good Average Poor VeryPoor
Datatransparency DataAccessibility Metadataquality
DataQuality InternaronalDimension
ARIADNED12.6FinalTestingReport
28
Comments:Thefeedbackfromrespondents inrelationtothegeneraluserequirementsofthe ARIADNE infrastructure reflects the results of the SUS scores. A higher level ofsatisfaction with the portal indicates that there are needs to be a more user-friendly,documented workflow for data ingest to the registry. Some responses indicate thatdepositorsdesiredahigherlevelofusercontrolovertheingestprocesssothatdatawouldappear according to provider’s needs in the portal. It is important to recognize that theportalrepresentsthefinalstageofthedataingestionworkflowandassuchismorelikelytoreflect satisfaction compared with earlier stages of content provision, which may haveinvolvedtechnicalandinfrastructuralissuestobeovercome,andisingenerallessrewardingintermsof“seeingtheresultsofworkdone”.
Additionalsurveythemes
Theonlinesurveyalsosoughttounderstandthe levelofsatisfactionwithotheraspectsofthe infrastructure including how familiar partners were with other features of theinfrastructure, how their data was ingested into the infrastructure and their level ofsatisfactionwiththeprocessfromingesttopublication.
ARIADNED12.6FinalTestingReport
29
TheARIADNECatalogueDataModel(ACDM)
ARIADNECatalogueDataModel is anextensionof theDataCatalogVocabulary (DCAT), aquasi-recommendation of the W3C Consortium that “is well-suited to representinggovernmentdatacataloguessuchasData.govanddata.gov.uk”.ThereasonforadoptingtheDCAT Vocabulary (apart from re-use) is that DCAT is proposed as a tool for publishingdatasets as Open Data. Its adoption places therefore ARIADNE in an ideal position forpublishing datasets as Open Data as well (see http://portal.ariadne-infrastructure.eu/about).
Respondents were asked about their level of familiarity with the ACDM as well as hownecessarytheyfeltaknowledgeoftheACDMwasfortheprovisionofcontenttoARIADNE.
Figure5.Chartshowingresponsestothequestion'HowfamiliarareyouwiththeACDM?'
The majority of respondents were familiar with the ACDM, the largest number (50%)considered themselves ‘somewhat familiar’. Respondents were further asked about hownecessarytheyfeltaknowledgeoftheACDMwasfor ingestingdata,withall respondentsconsidering iteithernecessary (70%)orpartiallynecessary (30%).60%ofrespondents feltthat there was insufficient training and resources available for using the ACDM withinARIADNE, while the remainder felt that there were sufficient resources. Severalrecommendationsfrompartnersonhowthismightbeachievedweresuggested,thesearegroupedaccordingtosimilaritybelow:
ARIADNED12.6FinalTestingReport
30
• ‘…Publishing worked examples of application…. I think more examples arenecessary….Addingmore examples ofmapping.Using amapping tool that showsthepreviewofthemappingresults………Byaddingtutorialsandexamplefiles(e.g.spreadsheetorxmlfile)……AguidelistinganddescribingeachACDMconcept(asitissomearemissing)…….’
• ‘….makeiteasiertofindthemostcurrentinformation/help/tutorials....TheHelplinkin the registry should be linking to the support portal http://support.ariadne-infrastructure.eu/. Within the support portal, there is a heading titled 'MappingGuidelines' but what follows aren't Mapping Guidelines', but rather the manyiterations of the Specification of the ARIADNE Catalogue Data Model. The actualmappingguidanceisburiedattheendofthesectionandcalledtheDataResourcesuserguide.Thiswholepageandthedocumentswithinitneedwillneedtobeeditedfor users not familiar with the ARIADNE project, and with their ease of use inmind…..’
Comments:ThespecificationfortheACDMiscurrentlycontainedinthe‘About’sectionoftheportal,[http://portal.ariadne-infrastructure.eu/about]theuserguideshouldbeupdatedwithmore examples and added to this section. The help section of the registry includesexampleExcelandXMLfiles.Theseshouldbeexpandedtoreflectmorevariedcontentandin particular the Excel file should include several more rows of sample data. This isparticularlyimportantgiventhattheseformatsweremostcommonlyusedbypartners(seebelow). In addition to this, as highlighted by a user above, the help link in the registryshould be linked to the support portal at http://support.ariadne-infrastructure.eu/#intro.Theexamplefilesshouldalsobeincludedonthisinterface.Ingeneral,however,itmustbenoted that understanding the ACDM and the required mappings requires substantialtechnical skills thatwere not always available for the content provider partners, as thosecommentingabove.Thusitislikelythatjustaddingmoreexamplesmaynotsolvetheissue,unlessaccompaniedbytrainingofcontentproviders.
IngestingData-TheARIADNEregistry
Thissectionshouldbeusedinordertoassessthebestmethodsforfuturedevelopmentoftheingestworkflowasitreflectstheformatsbestsuitedformappingfromprovider’snativedata(andpresumablythearchaeologicalcommunityatlarge).
ARIADNED12.6FinalTestingReport
31
Figure6.Chartindicatingwhichmethodsusedbypartnerstoingestdataintotheinfrastructure.Notethatsomepartnersusedmorethanonemethod.
ARIADNED12.6FinalTestingReport
32
Figure7.ChartindicatingthetypeofdigitalresourcesingestedtotheARIADNEinfrastructure.
Comments:ItshouldbenotedthatnopartnersusedtheRESTAPItoingestrecords.Thereistechnical documentation for the API contained in the Registry help section as well as inD12.3. It may be useful to edit this to provide more suitable documentation for non-technicalusers,withstepbystep instructions.SmallerproviderswithoutOAI-PMHtargetsarelikelytousetheExcelandXMLfilestoingestdata,thereforethecommentsnotedabove(aboutmoredetailedexamplefiles)shouldbeaddressedfornewusers,possiblysettingupatrainingprogram.Itwasrecommendedthatawebpagewithcleardetailsonhowdatacanbeadded to the registry is added to theportal (as thiswill generallybe the firstpointofinteractionwith the infrastructure). This page should include contact details for technicalpartnersforfuturecollaboratorswhowishtosubmitcontenttotheARIADNEinfrastructure.
ARIADNED12.6FinalTestingReport
33
The type of content ingested is given in Figure 6. Although no respondents reportedingestingGISdata,40%indicatedthattheywouldbelikelytoingestGISdatainthefuture.
Partners were given the opportunity to comment on any issues which they encounteredwhen ingesting data to the registry. Some reported that there were difficulties withaligning/normalising their native data with the requirements of the registry. Tworespondentscommentedontheneedforguidanceonwhetherdatashouldbeingestedatitemorcollection level.Thesedifficultiesweremainly resolvedbycontacting thepartnersresponsiblefortechnicalsupport.Respondentsindicatedagenerallevelofsatisfactionwithindividual technical support, how this supportwill bemaintained in the future should beconsideredbytheconsortiumatlarge.
MOReaggregationtool
OnlyasmallnumberofrespondentsreportedusingtheMOReaggregation,validationandenrichment tool. Users are directed to read section 4.3.2 which details internal testingcarriedoutonthisfeature.Ofthoserespondentswhousedthetool,allweresatisfiedwiththe enrichment of their data with only two enrichment services recorded as being used,LoGeoandThesaurimappings.Therefore,itisrecommendedinthefirstinstancethatthesetwo services, as well as PeriodO, be comprehensively documented in the tool and linksupdated to refer to ARIADNE documentation rather than LoCloud documentation. Someenrichmentmicro-servicesmaybe imrpovedwithmorecomprehensivedescriptions.ForadetailedworkflowofthistoolseeAnnexII.
Support/Helpdocumentation
There was an overall level of satisfaction with the support documentation with 70% ofrespondents finding the documentation either helpful or sometimes helpful. The chartbelowgivesdetailsonhowrespondentsfeltthedocumentationcouldbeimproved.
ARIADNED12.6FinalTestingReport
34
Figure8.Chartindicatinghowthesupportandhelpcouldbeimproved
Reiteratingwhatwasoutlined for the ingestprocess (seeabove) respondentscommentedthatexampledatasetsandmoremappingexampleswouldimproveuserexperience.
ARIADNED12.6FinalTestingReport
35
5 Conclusions
5.1 Registrytooltesting
Basedupon the requestsof theuser testing inD12.4 the resolutionofoutstanding issueshadbeencompletedinseveralareaswith50%ofissuesreachingsolutionnowand50%ofissues postponed to the next version of the system, where future new users will find afriendlierenvironment.This isparticularly thecase in the supportportal thatwillneed totakeintoaccountthecharacteristicsofanewwaveofcontentproviders.
Overall the user feedback from the new round of internal testing of the Registry Toolresultedinasatisfactoryscorewithspecificimprovementstobeplannedintoseveralareastoimprovetheusability,including:
• AdditionofUndo/Redofunction• Ensureallhelptextsarepresentincludinghelplinkswhichdescribethedifferent
propertiesoftheACDM• Employaclearervisualstructurefortheinterfacewhichreinforcesgroupingofkey
resourcingclasses
Areaswheretheregistrationtoolcouldbeimprovedincludealso:• Developmentofa“basic”and“expert/advanced”setting• Provisiontheuserfeedbackontheresultsoftheiractionanderrorreporting
Based upon the SUS testing average score result of 58.75 the Registry tool would beconsidered below average so some level of improvement is required in futuredevelopments.Thealreadynotedfactthatthecurrentproviders/testers/respondershadtopay a price for contributing to a system being developed and tested, with the relatedfrustration and dissatisfaction of availing, at least in the beginning, of a still incompletesystem,somehowmitigatesthisresult.
BasedupontheonlineusertestingtheRegistryreceivedpredominantlygoodratingacrossall the areas of testing including: data transparency, data accessibility, metadata quality,dataqualityandinternationaldimension.
Based upon the online testing users’ feedback effort should be made to improve theunderstanding and comprehension of the ACDM as familiaritywas relatively low and theprovisionoftrainingresourcesshouldbeconsideredasveryimportant.
ARIADNED12.6FinalTestingReport
36
5.2 MOReaggregationtooltesting
OveralltheuserfeedbackfromtheinternaltestingoftheMOReaggregationtoolresultedinamoderatetogoodscorewithgoodmarksgiventoallthetestingcriteria.Wherethereisthe potential for a slight improvement would be to update support documentation tocoordinatetooltoreferencetheARIADNEproject.BasedupontheonlineuserfeedbacktheMOReaggregationtoolwasnotbeingutilizedbyseveralprojecttestersandthebenefitsofitsuseshouldbepromotedwidely,specificallyasregardstheuseofthePeriodO,LoGeoandThesaurimappingfunctions.
5.3 Portaltesting
OveralltheuserfeedbackfromtheinternaltestingofthePortalresultedingoodscorewithgoodmarks given to all the testing criteria. The areas highlighted for slight improvementinclude:
• Abilityforuserstosaveanduseprevioussearchqueries• Improveerrorreportingfunctionsandvisibilityofhelpandguides
The first feature was not included in the initial specifications of the portal, and will beconsidered inARIADNE2.0.There isalwaysspace to improvehelp,andusers’ feedback isthebestway tounderstandwhichaspectsneed furtherdetail, as they look simple to thedeveloperbut are insteaddifficult for theuser,whileothers treatedwith greatdetail aremaybeoflittlerelevancefortheuser.
Based upon the SUS testing average score result of 77.5 the portal the Portal would beconsideredwellaboveaveragewithanadjectiveratingofgood.
Basedupontheonlineusertestingtheportalreceivedpredominantlygoodratingacrosstheareasof testing including:metadataquality,dataqualityand internationaldimension,andexcellenttogoodratingsfordatatransparencyanddataaccessibility
ARIADNED12.6FinalTestingReport
37
6 References
Brooke, J. (1996). SUS: A "quick and dirty" usability scale. In P. W. Jordan, B. Thomas, B. A.Weerdmeester,&A.L.McClelland(Eds.),UsabilityEvaluationinIndustryBangor, A., Kortum, P., & Miller, J.A. (2008) ‘The System Usability Scale (SUS): An EmpiricalEvaluation’,InternationalJournalofHuman-ComputerInteraction24(6).
Gavrillis, D., Papatheodorou, C., Canstantopoulos, P.,D12.2: Infrastructure Design, February 2015,http://www.ariadne-infrastructure.eu/Resources/D12.2-Infrastructure-Design
Gavrillis, D., Papatheodorou, C., Afiontzi, E., Makri, D.N., Dallas, C., D12.3: Initial InfrastructureImplementation Report, June 2015, http://www.ariadne-infrastructure.eu/Resources/D12.3-Initial-Infrastructure-Implementation-ReportHolzinger,A.,‘UsabilityEngineeringMethodsforSoftwareDevelopers’inCommunicationsoftheACMJan2005,Vol.48,No.1.Sauro,J.10ThingsToKnowAboutTheSystemUsabilityScale(SUS)availableathttp://www.measuringu.com/blog/10-things-SUS.php]
Selhofer, H., Geser, G., D2.1: First Report on Users’ Needs, April 2014, http://www.ariadne-infrastructure.eu/Resources/D2.1-First-report-on-users-needs
Selhofer,H.,Geser,G.,D2.2: SecondReport onUsers’Needs, February2015, http://www.ariadne-infrastructure.eu/content/view/full/1188Aframeworkofguidanceforbuildinggooddigitalcollections(2007)[http://www.niso.org/publications/rp/framework3.pdf]
ARIADNED12.6FinalTestingReport
38
ARIADNED12.6FinalTestingReport
39
7 Annex1:OnlineSurvey
Q1.AreyouaffiliatedwithanARIADNEparticipatingpartner?
AnswerChoices Responses
Yes 92.3%
No 7.7%
Total 13
Q2.Whatisthenameofyourorganisation?
Responses
SvenskNationellDatatjänst
AustrianAcademyofSciences
StuartEveConsulting
NationalInstituteofArchaeologywithMuseumattheBulgarianAcademyofSciences(NIAM-BAS)
ArheoVestAssociation
ArchaeologyDataService
GermanArchaeologicalInstitute
HungarianNatioanlMuseum
CentralInstitutefortheUnionCatalogueofItalianLibrariesandBibliographicInformation
ZRCSAZU
DigitalCurationUnit,AthenaResearchCentre
ARUP_CAS_17
Inrap
ARIADNED12.6FinalTestingReport
40
Q3.Whatbestdescribesyourrole?
AnswerChoices Responses No.
Archaeologist 76.9% 10
Researcher 69.2% 9
ComputerScientist 0% 0
Softwaredeveloper 0% 0
DigitalArchivist/RecordsManager/Informationscientist
38.5% 5
Other(pleasespecify)“DataManager”
7.7% 1
Q4.HowfamiliarareyouwiththeAriadneCatalogueDatasetModel(ACDM)?
AnswerChoices Responses No.
Veryfamiliar 10% 1
Quitefamiliar 30% 3
Somewhatfamiliar 50% 5
Notfamiliar 10% 1
Total 10
Q5.HownecessarydoyouthinkaknowledgeoftheACDMiswhenusingtheMetadataRegistry?
AnswerChoices Responses No.
Necessary 70% 7
Partiallynecessary 30% 3
Notnecessary 0% 0
Total 10
ARIADNED12.6FinalTestingReport
41
Q6.DoyoufeelthatuserswhoarenotfamiliarwiththeACDMwouldbeabletoingestmaterialintotheregistry,asitiscurrentlydesigned?
AnswerChoices Responses No.
Yes 20% 2
No 40% 4
Perhaps 40% 4
Total 10
Q7. Do you think there is sufficient training and resources (e.g. information/help/tutorials) onusingtheACDMwithinARIADNE?
AnswerChoices Responses No.
Yes 40% 4
No 60% 6
Total 10
ARIADNED12.6FinalTestingReport
42
Q8.HowcouldthetrainingandunderstandingoftheACDMbeimprovedinARIADNE?
Responses
Makeiteasiertofindthemostcurrentinformation/help/tutorials.
Publishingworkedexamplesofapplication.
Ithinkmoreexamplesarenecessary
TheHelplinkintheregistryshouldbelinkingtothesupportportalhttp://support.ariadne-infrastructure.eu/.Withinthesupportportal,thereisaheadingtitled'MappingGuidelines'but
whatfollowsaren'tMappingGuidelines',butratherthemanyiterationsoftheSpecificationoftheARIADNECatalogueDataModel.TheactualmappingguidanceisburiedattheendofthesectionandcalledtheDataResourcesuserguide.ThiswholepageandthedocumentswithinitneedwillneedtobeeditedforusersnotfamiliarwiththeARIADNEproject,andwiththeireaseofusein
mind.
Forarchaeologists,whoneverdonesuchthingitisdifficulttounderstandandperphapsastep-by-stepexplanationisneeded.
AddingmoreexamplesofmappingUsingamappingtoolthatshowsthepreviewofthemappingresults
AguidelistinganddescribingeachACDMconcept(asitissomearemissing)
Byaddingtutorialsandexamplefiles(e.g.spreadsheetorxmlfile)
Q9.Whichmethoddidyouusetoingestdata?Morethanoneoptionispossible.
AnswerChoices Responses No.
ARIADNEregistrywebtool(manually) 33.3% 3
Excelfile 33.3% 3
XMLfile 44.4% 4
RESTAPI 0% 0
OAI-PMH 33.3% 3
Other(pleasespecify) 0% 0
Total 9
ARIADNED12.6FinalTestingReport
43
Q10.Whattypeofdatadidyouingestintheregistry?Morethanoneoptionispossible.
AnswerChoices Responses No.
Collection 55.6% 5
Dataset 66.7% 6
Database 55.6% 5
GIS 0% 0
MetadataSchema 0% 0
Service 11.1% 1
Vocabulary 22.2% 2
TextualDocument 11.1% 1
License 0% 0
Total 9
ARIADNED12.6FinalTestingReport
44
Q11.Whattypeofdataareyoulikelytoingestinthefuture?
AnswerChoices Responses No.
Collection 66.7% 6
Dataset 66.7% 6
Database 55.6% 5
GIS 44.4% 4
MetadataSchema 11.1% 1
Service 0% 0
Vocabulary 22.2% 2
TextualDocument 11.1% 1
License 0% 0
Total 9
Q12.Didyouingestdataatitemorcollectionlevel?
AnswerChoices Responses
Itemlevel 22.2% 2
Collectionlevel 33.4% 3
Both 22.2% 2
Notapplicable 22.2% 2
Total 9
ARIADNED12.6FinalTestingReport
45
Q13.Didyouencounteranyproblemswheningestingdata?Pleasedetailconcisely.
Responses
None,ingestingtotheregistry.ThoughencounteredseveralproblemsingestingtoMoRE.Therewassomeconfusionastowhetherweshouldingestatcollectionoritemlevel,whichwefeel
shouldhavehadmoreguidanceon.
Asourdatawasthetestdata,therewerealotofproblems,butthiswastobeexpected.Weencounteredproblemswiththesizeofourfiles(harvestedfromtheOAI-PMHtarget)andwith
mappings(astheyhadtobedonemanuallybyATHENA).I'msuretheseissueswerelikelyresolvedbythetimeotherpartnerswereingestingtheirdata.
Wearestillworkingoningestingdata.
InourdatatheACDM:subjectcorrespondstodc:title,inwhichthevaluescanbeoneofthefollowingones:type(eg.Vagoconcapocchiaconombrellino)orobjectdefinition(eg.Fibula)orsubjecttitle(eg.ToroFarnese)orthenameoftheobject(eg.VasodiBaratti).Tonormalizethesevalueswehadtomapdc.titlevaluetothevocalularyAATandwecreatedaTAGnative:subjectwith
thenormalizedvalues.
Ingestingatitemlevelneedsanon-manualmethod.Excelisalwaysmentionedasapossibilitybutitturnedoutthatitwasnotso.Weneededexternalhelp(providedwithinARIADNE,manythankstoAchille!)tocreateXML'sfromExcel.Theprocesswasnottransparentandtherewasalotofback
andforth(duetothemistakesinourDB,butonesthatarecommoninmostDB's).AmoretransparentandaboveallsystemicwaytoingestExcelisneededsincethisistheonlymeansthat
90%ofarchaeologistsarefamiliarwith.IfthereisnoExceltheARIADNEportalwillbelimitedtothebiginstitutionswithinternalcomputertechnicians/scientistsupport.
Distributionofmultiplesubjectsandkeywords;placenameshavenotbeenusefulforgeocoding
ARIADNED12.6FinalTestingReport
46
Q14.PleaseratehowyoufeeltheregistryaddressesthegeneraluserequirementsoftheARIADNEinfrastructure:
AnswerChoices Excellent Good Average Poor Verypoor Total
Datatransparency[i.e.doestheregistryfacilitatetheaimofhavingabetteroverviewof
availabledata?]
0%
0
77.8%
7
22.2%
2
0%
0
0%
0
9
DataAccessibility[i.e.doestheregistryfacilitatedatabeingavailabletotheuserinanuncomplicatedway?]
11.1%
1
55.6%
5
22.2%
2
0%
0
11.1%
1
9
MetadataQuality[i.e.Doestheregistryensurethatingesteddataiswell-describedandimplementsstandardsfor
interoperability?]
11.1%
1
55.6%
5
33.3%
3
0%
0
0%
0
9
DataQuality(i.e.Doestheregistryensurethatingesteddataiscompleteandwell
organised?)
11.1%
1
77.8%
7
11.1%
1
0%
0
0%
0
9
InternationalDimension(Doestheregistryallowforeasy
accesstointernationaldata,e.g.throughdataintegration)
11.1%
1
66.7%
6
11.1%
1
11.1%
1
0%
0
9
ARIADNED12.6FinalTestingReport
47
Q15.HowwouldyourateyouroverallexperienceoftheARIADNEmetadataregistry?
AnswerChoices Stronglydisagree
Disagree Neutral Agree Stronglyagree
Total
IthinkthatIwouldliketousetheregistryfrequently.
11.1%
1
11.1%
1
22.2%
2
55.6%
5
0%
0
9
Ifoundtheregistryunnecessarilycomplex.
22.2%
2
33.3%
3
44.4%
4
0%
0
0%
0
9
Ithoughttheregistrywaseasytouse.
0%
0
11.1%
1
55.6%
5
33.3%
3
0%
0
9
IthinkthatIwouldneedthesupportofatechnicalpersontobeabletousetheregistry.
11.1%
1
44.4%
4
33.3%
3
11.1%
1
0%
0
9
Ifoundthevariousfunctionsintheregistrywerewell
integrated.
0%
0
0%
0
55.6%
5
44.4%
4
0%
0
9
Ithoughttherewastoomuchinconsistencyintheregistry.
0%
0
44.4%
4
33.3%
3
22.2%
2
0%
0
9
Iwouldimaginethatmostpeoplewouldlearntousethe
registryveryquickly.
0%
0
22.2%
2
33.3%
3
33.3%
3
11.1%
1
9
Ifoundtheregistryverycumbersometouse.
11.1%
1
44.4%
4
22.2%
2
22.2%
2
0%
0
9
Ifeltveryconfidentusingtheregistry.
11.1%
1
11.1%
1
44.4%
4
33.3%
3
0%
0
9
IneededtolearnalotofthingsbeforeIcouldgetgoing
withtheregistry.
11.1%
1
11.1%
1
44.4%
4
33.3%
3
0%
0
9
ARIADNED12.6FinalTestingReport
48
Q16.DidyouusetheARIADNEmetadataregistrywebtool?
AnswerChoices Responses No.
Yes 55.6% 5
No 44.4% 4
Total 9
Q17.Whattypeofaccountdidyouregisterwithinthetool?
AnswerChoices Responses No.
Individual 33.3% 2
Institution 66.7% 4
Other(pleasespecify) 0% 0
Total 6
Q18.Didotherindividualsfromyourinstitutionregisterseparately?
AnswerChoices Responses No.
Yes 16.7% 1
No 83.3% 5
Idon'tknow 0% 0
Total 6
Q19.Ifyes,didthisresultindifficultieswheningestingdatatotheregistry?
AnswerChoices Responses No.
Yes 0% 0
No 100% 4
Total 4
ARIADNED12.6FinalTestingReport
49
Q20.Doyoufeeltheregistrationprocedurecouldbeimproved?
AnswerChoices Responses No.
Yes 20% 1
No 80% 4
Total 5
Q21.Ifyes,pleaseselectfrombelow
AnswerChoices Responses No.
Tightercontrolsoverdataproviderse.g.registrationofinstitutionswhocontrollocalaccess
100% 1
Implementationofahierarchyofusers,e.g.institutionalmanager/editor
100% 1
Onfirstregistrationusersarepresentedwithawalkthroughvideo/navigationpop-ups/furtherguidanceetc
100% 1
Visibilityofsupportdocumentationimprovedonregistrylandingpage
100% 1
Other(pleasespecify) 100% 0
Q22.Howwouldyouratetheuserinterface/designofthemetadataregistrywebtool?
AnswerChoices Responses No.
VeryGood 0.00% 0
Good 60.00% 3
Average 40.00% 2
Poor 0.00% 0
Verypoor 0.00% 0
Total 5
ARIADNED12.6FinalTestingReport
50
Q23.Whatimprovementswouldyouconsidernecessarytothemetadataregistrywebtool?
AnswerChoices Responses No.
Betterlayout&design 20% 1
Improvedintroductoryand/orexplanatorydocumentation
60% 3
Improvedhelp/supportdocumentation 60% 3
Improvedtechnicalsupport 20% 1
Other(pleasespecify)“Thepossibilitytoloaddatadirectlyfrom
spreadsheetsorXMLfiles”
20% 1
Total 5
Q24.Pleaseratehowimportantitistoimprovethefollowingintheregistrywebtool:
AnswerChoices VeryImportant
Important Lessimportant
NotImportant
No.
Design/UserInterface 0%
0
0%
0
60%
3
40%
2
5
Supportdocumentation 40%
2
60%
3
0%
0
0%
0
5
Facilityforpreviewingdata 40%
2
20%
1
40%
2
0%
0
5
Helptexts/pop-ups 20%
1
40%
2
20%
1
20%
1
5
Validation/Qualitycontrols 60%
3
40%
2
0%
0
0%
0
5
ARIADNED12.6FinalTestingReport
51
Q25.DidyouusetheMOReaggregationtoolaspartofyouringestprocess?
AnswerChoices Responses No.
Yes 33.3% 3
No 33.3% 3
Idon'tknowwhatthisis 33.3% 3
Total 9
Q26.Wereyou…
AnswerChoices Responses No.
familiarwiththeMORetoolfromotherEUprojects,e.g.Locloud,Carare
33.3% 3
giventrainingintheMORetoolaspartoftheARIADNEproject 0% 0
learnedhowtousetheMORetoolthroughonlinesupport/documentation
33.3% 3
Other(pleasespecify)“Idon'tknowwhetherthiswasunusualtoARIADNEpartners,butourMORetoolwassetup,andcompletedinthefirstinstancebytheDCU.
Itwasonlymuchlaterthatweweretoldabouttheaccount”
33.3% 3
Total 9
ARIADNED12.6FinalTestingReport
52
Q27.DidyouuseanyenrichmentservicesinMORe?
AnswerChoices Responses No.
Yes 66.7% 2
No 33.3% 1
Total 3
Q28.Ifyes,whichenrichmentservicesdidyouuse?
AnswerChoices Responses No.
Language 66.7% 2
DaiGazeteer 0% 0
LoGeo 50% 1
Geocode 0% 0
Rev.Geocode 0% 0
GeoInvertion 0% 0
Pelagios 0% 0
Vocabulary 0% 0
BackLink 0% 0
Thesaurimappings 100% 2
PleiadesPlus 0% 0
Perio.domappings 0% 0
Other(pleasespecify) 0% 0
BackLink 0% 0
ARIADNED12.6FinalTestingReport
53
Q29.Wereyousatisfiedwithyourenricheddata?
AnswerChoices Responses No.
Yes 100% 2
No 0% 0
Total 2
Q30.Didyouuseanyofthehelp/supportdocumentationfortheregistry?
AnswerChoices Responses No.
Yes 75.00% 6
No 25.00% 2
Total 8
Q31.Didyoufindthisdocumentationhelpful?
AnswerChoices Responses No.
Yes 14.3% 1
Sometimes 57.1% 4
No(Pleasegivedetails)“Wedidn'thavetoconsultthisdocumentastheingestionhadbeen
doneforusinthefirstinstancebytheDCU.”“It'stoobriefandincomplete”
28.6% 2
ARIADNED12.6FinalTestingReport
54
Q32.Howdoyouthinksupport/helpdocumentationcouldbeimproved?
AnswerChoices Responses No.
Morevisibleand/oraccessible 66.7% 4
Moredetailed 50% 3
Lessdetailed 0% 0
Needstobeupdated 50% 3
Needsmoreoptionse.g.livetechsupport,walkthroughvideosetc. 50% 3
Other(pleasespecify)“Exampledatasetstogetanideaofwhatshouldbeenteredwhere.”
“moremappingexamples”
33.3% 2
Q33.PleaseusethisopportunitytocommentonanyotheraspectsoftheRegistrywhichyoufeelneedimprovement:
AnswerChoices
BettercommunicationofwhatourresponsibilitieswerewithregardtoMORe.Aworkshopperhaps,aswasdonewiththeRegistry.PoordesignwithregardstomappingtotheGettyAAT.AIAChaddevelopedourOAI-PMHtodelivertheAATURIs,butMORerequiredustolookuptheAATpreferredterms,whichwebelieveresultedinunnecessarydoublehandlingofthedatato
reachthesameendresult.
Thereweresomeproblemswiththeacdm:temporal
Ichoseaverageforquestions14-15wedidn'tusetheregistry,andtherewasnowaytoskipthequestions.
ARIADNED12.6FinalTestingReport
55
Q34.WereyousatisfiedwithhowyourdataappearedintheARIADNEportalafteringestion?
AnswerChoices Responses No.
Yes 55.6% 5
No(pleasespecify)“Ithasnotbeeningestedintotheportalyet.”
“Wehadseveralproblemstobeginwith.Firstourdatespansweren'tworking,second,weweren'tadvisedontheneedtohavelandingpagesastheywereinitiallyindicatedasoptional.We'redissatisfiedwiththefactthattheuserisunabletoselecttwopublishersinthefilteroptions,sothatour
datacanbecomparedwithothers.”“Wearestillworkingoningestingourdataintheportal”
“Placenameshavenotbeenusefulforgeocodinganditstilldoesn'tappearsthemappingofthesauri(enrichment)”
44.4% 4
Total 9
Q35.PleaseratehowyoufeeltheARIADNEPortaladdressesthegeneraluserequirementsoftheARIADNEinfrastructure:
AnswerChoices Excellent Good Average Poor Verypoor Total
Datatransparency[i.edoestheportalfacilitatetheaimofhavingabetter
overviewofavailabledata?]
33.3%
3
66.7%
6
0%
0
0%
0
0%
0
9
DataAccessibility[i.e.doestheportalfacilitatedatabeingavailabletotheuserinanuncomplicatedway?]
33.3%
3
66.7%
6
0%
0
0%
0
0%
0
9
MetadataQuality[i.e.Doestheportalensurethatdataiswell-describedand
implementsstandardsforinteroperability?]
0%
0
77.8%
7
22.2%
2
0%
0
0%
0
9
DataQuality[i.e.Doestheportalensurethatdataiscompleteandwell
organised?]
0%
0
88.9%
8
11.1%
1
0%
0
0%
0
9
InternationalDimension[i.e.Doestheportalallowforeasyaccessto
internationaldata,e.g.throughdataintegration]
55.6%
5
44.4%
4
0%
0
0%
0
0%
0
9
ARIADNED12.6FinalTestingReport
56
Q36.HowwouldyourateyouroverallexperienceoftheARIADNEportal?
AnswerChoices Stronglydisagree
Disagree Neutral Agree Stronglyagree
Total
IthinkthatIwouldliketousetheportalfrequently
0%
0
0%
0
22.2%
2
55.6%
5
22.2%
2
9
Ifoundtheportalunnecessarilycomplex.
44.4%
4
44.4%
4
0%
0
11.1%
1
0%
0
9
Ithoughttheportalwaseasytouse. 0%
0
0%
0
22.2%
2
55.6%
5
22.2%
2
9
IthinkthatIwouldneedthesupportofatechnicalpersontobe
abletousetheportal.
66.7%
6
22.2%
2
0%
0
11.1%
1
0%
0
9
Ifoundthevariousfunctionsintheportalwerewellintegrated.
0%
0
0%
0
33.3%
3
55.6%
5
11.1%
1
9
Ithoughttherewastoomuchinconsistencyintheportal.
33.3%
3
55.6%
5
0%
0
11.1%
1
0%
0
9
Iwouldimaginethatmostpeoplewouldlearntousetheportalvery
quickly.
0%
0
0%
0
11.1%
1
55.6%
5
33.3%
3
9
Ifoundtheportalverycumbersometouse.
44.4%
4
33.3%
3
0%
0
22.2%
2
0%
0
9
Ifeltveryconfidentusingtheportal. 0%
0
0%
0
22.2%
2
55.6%
5
22.2%
2
9
IneededtolearnalotofthingsbeforeIcouldgetgoingwiththe
portal.
44.4%
4
33.3%
3
22.2%
2
0%
0
0%
0
9
ARIADNED12.6FinalTestingReport
57
Q37.Areyousatisfiedwiththedesign/layout/userinterfaceoftheARIADNEportal?
AnswerChoices Responses No.
Yes 88.9% 8
No(pleasespecify)“WhenIsearchforsomethingitisnotclearwhetherthereisdocumentationtobedownloadedornot.InmanycaseswhenIfoundinterestingthingsIrealizedthatthereisnofurtherdocumentationthatIcandownloadonlyalittleinformationonthesite.Sincetherearelotsofsitesinthedatabasetoomuchtimeisneededtofinddocumentationssinceyouhavetoclickthrough
dozensofthemtofinddocumentations.”
11.1% 1
Total 9
ARIADNED12.6FinalTestingReport
58
Q38.Whichofthefollowingfacets/filtersdoyoufeelareimportant?
AnswerChoices VeryImportant
Important SlightlyImportant
Neutral NotImportant
Total
Where 100%
9
0%
0
0%
0
0%
0
0%
0
9
When 77.8%
7
22.2%
2
0%
0
0%
0
0%
0
9
ResourceType 55.6%
5
33.3%
3
11.1%
1
0%
0
0%
0
9
NativeSubject 33.3%
3
33.3%
3
33.3%
3
0%
0
0%
0
9
DerivedSubject 33.3%
3
33.3%
3
33.3%
3
0%
0
0%
0
9
Keyword 55.6%
5
22.2%
2
22.2%
2
0%
0
0%
0
9
Contributor 22.2%
2
44.4%
4
22.2%
2
11.1%
1
0%
0
9
Publisher 22.2%
2
33.3%
3
22.2%
2
11.1%
1
11.1%
1
9
Place 77.8%
7
0%
0
22.2%
2
0%
0
0%
0
9
Dating 88.9%
8
0%
0
11.1%
1
0%
0
0%
0
9
Rights 44.4%
4
22.2%
2
33.3%
3
0%
0
0%
0
9
Language 22.2%
2
44.4%
4
33.3%
3
0%
0
0%
0
9
ARIADNED12.6FinalTestingReport
59
Q39.Doyoufeelthatthemetadataintheportalis..
AnswerChoices Yes Partially No Total WeightedAverage
ofasufficientstandard 77.8%
7
22.2%
2
0%
0
9 1.22
isdetailedenough 66.7%
6
33.3%
3
0%
0
9 1.33
displayscorrectly 66.7%
6
22.2%
2
11.1%
1
9 1.44
iseasilyunderstandable 44.4%
4
55.6%
5
0%
0
9 1.56
Q40.Doyouthinkanyadditionalmetadatafieldsshouldbedisplayedintheportal?
AnswerChoices Yes Total
No 88.9% 8
Yes(pleasespecify)“Itwouldbeusefultoshowwherearedocumentationsthatcanbe
downloaded”
11.1% 1
Q41.Areyousatisfiedwiththebrowsefunctions(Where,When,What)oftheportal?
AnswerChoices Yes Total
Yes 66.7% 6
No(Pleasegivedetails)“Largelyyes,butIbelievethe'DisplayasSearch'buttonthatfeaturesontheWherebrowser,needstobemademoreprominentontheWhenbrowser.”“IftheARIADNEsubjectisthesameasAATsubject,itwouldbebettertosay
thatandhavethesubjectsbetransparent.”“themapjumpswhenIzoomonit(zoomsout),itcannotbeusedproperly”
33.3% 3
ARIADNED12.6FinalTestingReport
60
Q42.Areyousatisfiedwiththesearchfunctionoftheportal?
AnswerChoices Yes Total
Yes 88.9% 8
No(Pleasegivedetails)“WedidnotyetexactlytestitbutcurrentlyIamstrugglingtofindthesites
thatIknowarethere(e.g.IronAgeburialmoundcemeteryinStična,Slovenia).”
11.1 1
Q43.Didyouhaveanyissuesaccessingdatathroughtheportal?
AnswerChoices Yes Total
No 100.00% 9
Yes(Pleasespecify) 0.00% 0