d6.1 dissemination and liaison plan - big data benchmarking€¦ · deliverable d6.1 dissemination...

DataBenchGrantAgreementNo78096

1

Evidence Based Big Data Benchmarking to Improve BusinessPerformance

D6.1DisseminationandLiaisonPlan

DeliverableD6.1DisseminationandLiaisonPlan


2

KeywordsBenchmarking, big data, big data technologies, business performance, economicindicator,Europeansignificance,industrialrelevance,performancemetrics,tool,usecases,dissemination,marketing,BigDataValuePPP,image.

DisclaimerThis document reflects the authors view only. The European Commission is notresponsibleforanyusethatmaybemadeoftheinformationthisdocumentcontains.

CopyrightNoticeCopyrightbelongs to theauthorsof thisdocument.Useof anymaterials from thisdocumentshouldbereferencedandisattheuser'sownrisk.

DeliverableD6.1 DisseminationandLiaisonPlan

Workpackage WP6

Task 6.1

Duedate 15/09/2018

Submissiondate 15

/09/2018

Deliverablelead ATOS

Version 2.0

Authors ATOS(NuriadeLama,AnaMorales)

IDC(CristinaPepato,StefaniaAguzzi)

Reviewers FrankfurtBigDataLab(TodorIvanov)

IDC(RichardStevens)



3

AbstractDataBenchhasthepotentialtobecomeakeyelementoftheBigDataValuePublic-Private Partnership, since it will contribute to measure the real impact of theinvestmentsmadebyindustryandtheEuropeanCommissioninthispartnership.Itwill support the benchmarking of Big Data Technologies, therefore helping tounderstand the progress with respect to the state of the art but will also relatetechnicalperformanceindicatorswithbusinessindicators,establishingabridgethathasneverbeenbuiltbefore.The challenges are however important. Among them,DataBenchwill require tightcollaborationwithanumberofplayersandcommunities.Workinghand-in-handwithbenchmarkingcommunitiesaswellaswithpilotsandusecases(ideallycomingfromPPPprojects) is key. Theywill be contributors but also validatorsof the so calledDataBenchToolBox,mainproducttobeproducedbytheproject.AlongthethreeyearsdurationofDataBenchwewillhavetogainthecredibilityandrecognition of those communities, since -in most cases- they will be the firstadopters/usersoftheDataBenchoutcomes.WP6will supportDataBench in creating the connections and engagingwith thosecommunities, will define and will implement a dissemination and communicationstrategyandwillpavethepathtowardstheself-sustainabilityoftheresultsbeyondtheprojectduration.ThisdeliverabledescribestheinitialstepsinthatendeavourbysharingtheDisseminationandLiaisonPlan.Thedocumentidentifiesspecifically:(1)target communities/audience, (2) Phases of the dissemination strategy, (3) tools,channelsandmechanismsthatwillbeusedand(4)KeyPerformanceIndicators.



4

TableofContentsAcronymsandAbbreviations...................................................................................................................6ExecutiveSummary.......................................................................................................................................71. IntroductionandObjectives............................................................................................................82. DataBenchConcept,ObjectivesandAssets...........................................................................103. TargetAudience.................................................................................................................................143.1 TheBigDataValuePPPframeworkandprojects....................................................143.2 InnovationSpacesorI-Spaces...........................................................................................173.3 OtherpotentialusersofDataBench................................................................................203.4 Benchmarkingcommunities..............................................................................................21

4. DataBenchtiming:DisseminationPhases.............................................................................234.1 Phase1-Createawareness(M1-M12).........................................................................234.2 Phase2-Increasethepotentialimpact(M13-M24)..............................................234.3 Phase3-MaximiseResults(M25-M34).......................................................................234.4 Phase4-Valorisation(M34-M36+M>36)...................................................................24

5. ImplementationoftheStrategy:toolsandchannels........................................................255.1 DataBenchbranding...............................................................................................................255.2 MarketingMaterial.................................................................................................................255.3 Strategyontheuseofwebchannels..............................................................................255.4 BDVAtoolsandengagementwiththePPPcommunity........................................275.5 Events............................................................................................................................................30

6. KPIsandMonitoringFramework..............................................................................................367. Conclusions...........................................................................................................................................378. References.............................................................................................................................................389. AnnexI.StateofplayofsectorsintheBigDataPPP........................................................39

Transport,mobilityandlogistics...............................................................................................39Bioeconomy..........................................................................................................................................41SmartManufacturingIndustry....................................................................................................44Healthcare.............................................................................................................................................46Telecommunications........................................................................................................................48Media.......................................................................................................................................................49



5

TableofFiguresFigure1DataBenchexpectedresults................................................................................................10Figure2FunctionalviewoftheDataBenchecosystem...........................................................12Figure3Portfolioanalysis:mappingwiththeBDVSRIA.........................................................15Figure4VisualrepresentationofthePPPprojectportfolio..................................................16Figure5DistributionofBigDataCentersofExcellence...........................................................20Figure6DIHwithcompetencesinBigData,dataminingandDBM...................................20Figure7SampleofNationalInitiativesonBigData...................................................................21Figure8ListofbenchmarkinginitiativesanalysedbyDataBench.....................................22Figure9DataBenchlogo..........................................................................................................................25Figure10DataBench@PPPportal......................................................................................................28Figure11TheBigDataLandscape......................................................................................................29Figure12ThePPPInnovationMarketplace...................................................................................29Figure13ThePPPWebinarsSeries...................................................................................................30Figure14BigDataPPPMeet-upSofia...............................................................................................31

TableofTablesTable1:Descriptionofi-SpacesaspotentialusersofDataBench.......................................19Table2:ListofpotentialeventsforDataBench............................................................................35Table3:DisseminationandcommunicationKPIs.......................................................................36



6

Acronyms and Abbreviations Acronym TitleAI ArtificialIntelligenceBDBC BigDataBenchmarkingCommunityBDT BigDataTechnologiesBDV BigDataValueBDVA BigDataValueAssociationBSS BusinessSupportSystemCMS ContentManagementSystemCoE CenterofExcellenceCRM CustomerRelationshipManagementDG DirectorateGeneralDIAS DataInformationAccessServicesDIH DigitalInnovationHubEBDVF EuropeanBigDataValueForumEC EuropeanCommissionECSO EuropeanCyberSecurityOrganisationEFFRA EuropeanFactoriesoftheFutureResearchAssociationENoLL EuropeanNetworkofLivingLabsEU EuropeanUnionGDP GrossDomesticProductGDPR GeneralDataProtectionRegulationHIPEAC HighPerformanceandEmbeddedArchitectureandCompilationHPC HighPerformanceComputingHTML HyperTextMarkupLanguageICT InformationandCommunicationTechnologiesIoT InternetofThingsJRC JointResearchCenterKPI KeyPerformanceIndicatorLDBC LinkedDataBenchmarkingCouncilNFV NetworkFunctionVirtualizationOSS OperationalSupportSystemPPP PublicPrivatePartnershipRoI ReturnonInvestmentsSRIA StrategicResearchandInnovationAgendaSC SteeringCommitteeSDIL SmartDataInnovationLabSDN SoftwareDefinedNetworksSME SmallandMediumEnterpriseTC TechnicalCommitteeTF TaskForceTFP TotalFactorofProductivityTRL TechnologyReadinessLevelUK UnitedKingdomUSA UnitedStatesofAmericaWP WorkPackage



7

Executive Summary WP6willsupportDataBenchincreatingconnectionsandengagingwiththerelevantcommunitiesandstakeholders,willdefineandwill implementadisseminationandcommunicationstrategyandwillpavethepathtowardstheself-sustainabilityoftheresultsbeyondtheprojectduration.ThisdeliverabledescribestheinitialstepsinthatendeavourbysharingtheDisseminationandLiaisonPlan.Thedocument identifiesspecifically: (1) target communities/audience, (2) Phases of the disseminationstrategy,(3)tools,channelsandmechanismsthatwillbeusedand(4)KPIs.Among the target communities it is worth mentioning the PPP project portfolio,benchmarking communities, I-Spaces (data-driven experimentation environments)andnetworksofresearchorinnovationstructuresthatcouldwellbecomepotentialusers of the DataBench ToolBox (e.g. Big Data CoE, DIH). All of them are brieflydescribedinthedocumentandthevaluepropositionisanalysed.



8

1. Introduction and Objectives Itisnotbychancethatweusethecriterionofimpactwhenweevaluateprojectsandinvestment opportunities. And the reason is that we all know that scientific andtechnical excellence is not a guarantee to be successful. Your system may workperfectlywellandreactasexpected,butyoumaybedoingsomethingwherethevalueisnotevidentforusers,orwherethevaluepropositionisnotrelevantenoughastoinvestinit.So,beingtechnicalperformanceapre-requisitebutnotaguaranteeforadoption, we have to create an environment that maximizes the probability ofconvincingusers. This is evenmore important in the case ofDataBench, since themajoroutcomeoftheprojectwillbeatoolthatrelatestechnicalperformanceofBigDataTechnologies(sometimesreferredtoasBDT)withbusinessindicators.Thetool,knownasDataBenchToolbox,isfullydependentonthecollaborationwithdifferentstakeholdersalongtheproductcycle,fromitsinceptiontillitsusage:

• Beforeitislaunched(designandimplementationphase)itwillbefedbyusecasesthathelpusunderstandrelationshipsbetweenindicatorsandthatallowthesystemtolearnandmaketherightrecommendations:thisimpliesworkingwithmany companies in diverse sectors andwith heterogeneous technicalenvironments;

• Before it is launched (design and implementation phase) it will requireintegrationof existingbenchmarks: theDataBenchToolboxwillnotreplacebenchmarksthatworkwellbutwillprovideanenvironmentthatfacilitatestheworkof thosepeoplethatrunthebenchmarkingprocesses(forexample,byincludingseveralbenchmarksinthesamesystem).Itisthereforeimportanttoworkcloselywiththebenchmarkingcommunity;

• Whenitislaunched,theDataBenchToolboxwillmakesenseonlyifthereisacommunityofusers ready toadopt themethodologyand tooldevelopedbyDataBench.

Therefore, developing a solid network of collaborators, contributors andadopters/usersisessentialtothesuccessofDataBench.Wewillhavetoworkongainingthecredibilityofthosestakeholders,startingwithsomeambassadorsandlateronusingnetworkstomultiplytheDataBencheffect.Onthe other hand, we should not neglect the impact of running a well-designedcommunicationcampaign.Communicationanddisseminationactionsbuiltontopofagoodproductwillforsurecontributetoachievetheexpectedimpact.WP6istheWPinchargeofConsensusBuilding,DisseminationandExploitation.Assuch,itwillhaveamajorresponsibilityinachievingthebiggestpossibleimpact.TheworkinthisWPisorganizedaroundthreemajorlines:

• Communityengagementoractiveinvolvementofstakeholders• Awareness,disseminationandmarketing&communicationcampaign• Exploitation,DataBenchadoptionandsustainability

D6.1providesthe initialversionof theDisseminationandLiaisonPlan,addressingelementsofthetwofirstactionlines.Theplanisnotaclosedandhermeticdocument.Onthecontrary,itwillberevisedonaperiodicbasis.Updatedversionsandpotentialchangesofdirectionwillbereportedinthefollow-updeliverables(D6.3andD6.4),whichwillalsocontainthereportofactivitiescarriedoutbyDataBenchaccordingto



9

thisstrategyaswellasacriticalassessmentoftheachievementsagainstthedefinedKey Performance Indicators (KPIs). This framework will undoubtedly feed D6.5ExploitationandSustainabilityplans.D6.5willaddressthethirdactionlineofthelistandwillpavethewaytotheself-sustainabilityofDataBench.ConnectionswiththeotherWPsoftheprojectareimportant,sincetheyprovidethecontent, aswell as themethodologies toworkwith thedifferent stakeholdersandcommunitiesoncetheyhavebeenreachedoutbyWP6.Wecouldnot finish this introductionwithoutreferring to the special environmentDataBenchbelongsto.DataBenchispartoftheBigDataValuePPP(alsoreferredtoasBDVPPPorPPPinthisdocument).TheprivatepartofthePPPistheBigDataValueAssociation(orBDVA),whichrepresentstheviewsofindustry;thepublicpartofthecontract is represented by the European Commission (EC). While planning,monitoringandsomecontent-relatedactivitiesarepartoftheactivitiesperformedbyBDVA,theimplementationoftheso-calledStrategicResearchandInnovationAgenda(SRIA)releasedbyBDVAhappensmainlythroughtheprojects.DataBenchisoneofthose projects and as part of the program, collaboration and cooperationwith itspeersisexpectedforacoherentdevelopmentofthePPP.FurthermoreDataBenchhasthepotentialtobecomeanessentialelementoftheprogramsincea)itcouldhelptoassesstheimpactofBDTinthedifferentpilots/usecases/environmentsofthePPPandb) thanksto the learningexperienceof thesystem, itcouldhelpcompaniestounderstandandtaketherightdecisionswhenitcomestodecidewhichtechnologies,frameworks or architectures are needed to fulfil their real needs and businessrequirements.



10

2. DataBench Concept, Objectives and Assets The goal of DataBench is to design a benchmarking process helping EuropeanorganizationsdevelopingBigDataTechnologiestoreachforexcellenceandconstantlyimprove their performance. This will be done by measuring their technologydevelopment activity against parameters of high business relevance. In order toachievethis,DataBenchhasdefinedthefollowingconcreteobjectives,asstatedinourinitialdisseminationmaterial:

1) ProvidetheBDTstakeholdercommunitieswithacomprehensiveframeworktointegratebusinessandtechnicalbenchmarkingapproachesforBDT.

2) Perform economic and market analysis to assess the “European economicsignificance”ofbenchmarkingtoolsandperformanceparameters.

3) EvaluatethebusinessimpactsofBDTbenchmarksofperformanceparametersofindustrialsignificance.

4) Develop a tool applying methodologies to determine optimal BDTbenchmarkingapproaches.

5) Evaluation of the DataBench Framework and Toolbox in representativeindustries, data experimentation/integration initiatives (ICT-14) and Large-ScalePilot(ICT-15).

6) LiaisecloselywiththeBDVA,ICT-14and15projectstobuildconsensusandtoreach out to key industrial communities, to ensure that benchmarkingrespondstorealneedsandproblems.

Thosestatementsrepresentthe“how”,i.e.,thewaytheprojectwillrealizeitsmajorgoal,butforthepurposeofcommunicatingtheproject,wewillfocusmuchmoreonthe “what”, i.e. the assets and outcomes thatwill be generated by the project andconstituteitsmainessence.Theywillbethetangibletoolsthatwillbeprovidedtothecommunity.

Figure1DataBenchexpectedresults

Eachoftheassetslistedinthepictureaboverelatestotheothersandaddsvaluetothe potential usage ofDataBench. However, and aswewill see later on, themainproductisthesocalledDataBenchToolbox.



11

Asdescribedin[1],theDataBenchToolboxwillallowuserstoselectfromtheplethoraofexistingbenchmarkstheone(s)thatsuitstheirneeds,deployanduseit,aswellasthepossibilitytouploadtheresultsofthebenchmarkexecutiontogetahomogenizedsetofmetrics,includingbusinessinsights.ForthepurposeofanoverallunderstandingoftheDataBenchecosystemweincludeherethefunctionalviewwithitselements,asdepictedin[1].Moredetailedtechnicaldescriptionscanbefoundinthatdocument.

• The DataBench Toolbox: The DataBench Toolbox is the core technicalcomponentoftheDataBenchFramework.ItwillbetheentrypointforusersthatwouldliketoperformBigDatabenchmarkingandwillultimatelydeliverrecommendations and business insights. It will include features to reuseexistingbigdatabenchmarks,andwillhelpuserstosearch,select,download,execute and get a set of homogenized results. TheToolbox is based on theanalysisandworkperformedinotherareasoftheproject.

• BenchmarksintegratedintotheToolbox:externalBigDatabenchmarksarethe input to the whole DataBench framework, and in particular to theDataBenchToolbox. These benchmarkswill bemade available to the usersthroughaneasy-to-useToolboxuserinterface.ThedegreeofintegrationoftheToolboxwiththeexistingbenchmarkswillvarydependingonthecases.

• BusinessKPIs:DerivingbusinessKPIs fromtheresultsofrunningBigDatabenchmarksisnotastraightforwardprocess.Itwilldependonthebusinesscontext and therefore it is not completely clear towhat extend it could beautomated. However, different approaches to derive business insights arecurrentlyunderstudyinthecontextofDataBenchandtheywillprovidethebasis for one of themost important added values of this initiative,with anextremely high potential impact for the Big Data Value PPP and the dataecosysteminEurope.

• AI framework: The framework will provide recommendations to thebenchmarking community based on past experiences. This componentwillalsobeintegratedwiththeToolbox.



12

Figure2FunctionalviewoftheDataBenchecosystem[1]

LookingdeeperintotheUniqueSellingPropositionsoftheDataBenchToolboxwecanderivetheoverallbenefitsproposedbyDataBenchandthewaytheyfitwiththeneedsof the market. Some simple messages that will be the basis of our initialcommunicationcampaignaredepictedhere:

• Organizationsneedmethodstoselect thebestBigDataTechnologies(BDT)fortheirbusinessneeds,particularlyinemergingapplicationareassuchastheIoTsmartenvironments,ortheAIandroboticsimplementations;

• There is still a wide gap between the capability to monitor technicalperformanceofBDTandtheassessmentoftheirbusinessimpacts,andthisisoneofthebarriersslowingdownadoption;

• There is a lack of objective, evidence-based methods measuring thecorrelationbetweenBDTtechnologybenchmarksandbusinessbenchmarksaswellasBDTcapability to impact thecompetitivenessandmarketsuccessoforganizations

• Noonehasyetbridgedthegapbetweentechnicalbenchmarksandeconomicperformance

In summary, there is a plethora of Big Data tools and frameworks, but how cancompaniesenteringtheBigData“era”knowwhichone(s)canfulfiltheirneedsandtheirbusiness/technicalenvironment?Sometimesatechnicalbenchmarkisdirectlyrelatedtoanunderstandablebusinessbenefit.Forexample,weallagreethatifyouhavetoprovidearesultofanoperationinafinancialenvironment(investment)andyourvelocity(numberofoperationsprocessedpersecond)increases,youwillbeableto react faster. In such environment reacting faster is a clear benefit, but bettertechnicalperformanceofsomeindicatorssometimesdoesnottranslatesoeasilyinto



13

abetterRoI.DataBenchwillhelptorelatetechnicalindicatorstobusinessones,andhence will provide the tools to better understand the economic impact of sometechnicaldecisionsassociatedtodifferentchoicesofBDT.For DataBench there are obvious challenges; ensuring that the benchmarksidentified respond to actual business needs is one of them and getting theacceptanceandrecognitionbytheindustrialcommunityisanotherone.Ifthefirstoneissuccessfullyimplemented,thenthepathtowardsthesecondonewillbeeased.Themainobjectiveof theworkproposed inthisDisseminationPlan isprecisely tosupportDataBenchinthosechallenges;tocreatetherelationships,mechanismsandtoolsthatwillhelpDataBenchtoworkhand-in-handwiththerightcommunities,theonesthata)willhelpDataBenchtocreatetherightproductand/orb)willbecomeusersorcustomersoftheDataBenchToolbox.TheextenttowhichtheDataBenchToolboxfulfilsitsfunction(i.e.providestherightrecommendationsandrelates technicalbenchmarkswithbusiness indicators inanaccurateway)willbethemajorindicatorof(technical)successoftheproject.The extent to which DataBench is adopted will be the main indicator of(business)successoftheproject.Exactlyasmentionedfewparagraphsabove,thetechnicalsuccessofDataBenchwillbe a pre-condition for the business success. However, the second will also bedependent on additional forces and elements. Here is where the communicationcampaigndesignedbytheprojectwillbeparticularlyrelevant.TheDisseminationstrategyincludesthreemajorelements,besidesthedefinitionofthe“product”:

• Identification of target communities (who is our customer/recipient),includingthevaluepropositionforthem(whyshouldtheybeinterestedaboutDataBench)

• Disseminationphases(when);understandingwhendifferentactionshavetobe implemented is important. For example, too muchmarketingwhen theproductisnotavailablecanleadtoasituationwherepotentialusersgettired.The oneswho are not retainedwill probablynot comeback.Messageswillobviouslyevolveoverthetimetoo,aswellasthetoolstobeusedandthewaytheywillbeused.

• Tools,mechanismsandchannels(howorwhere):therangeofchannelsthatwill be used by DataBench include web tools (website, social networks),disseminationmaterial,eventsorpublications, tonamesomeof them.Theywill also evolve over the time from different points of view (message,frequencyofuse).



14

3. Target Audience

3.1 The Big Data Value PPP framework and projects

InOctober of 2014 the EuropeanCommission and theBigDataValueAssociation(fromnowonreferredtoasBDVA[2])signedacontractualagreementforaPublicPrivatePartnershipfocusedonBigDataTechnologies,or,astheformalnamesays,onBig Data Value (i.e. the ability to transform data into value thanks to BDT). Thecontractualagreementincludesawideumbrellaofobjectives,butsomeofthem-thatwelistbelow-arespecificallyrelevantinthecontextoftheworkDataBenchproposes[3]:Objectivesforimprovedcompetitiveness:

• Develop solutions leading towards the use of Big Data Technology forincreased productivity, optimized production, more efficient logistics, andeffectiveserviceprovisionfrompublicandprivateorganizations,

• Developanddiffuseabetterunderstandingofbusinessopportunitiesof theBigDatasector,

• Drive the take-up and integration of Big data value services in private andpublicdecision-makingsystems,

Innovationobjectives:

• Validatetechnologiesfromatechnicalandbusinessperspectivethroughearlytrials in cross-organizational, cross-sector and cross-lingual innovationenvironments.

Aswe can see, the PPPplays a very important role in driving the use of BigDataTechnologies,orsaidinotherwords,indrivingthedemandside.Oneofthereasonsthat isusuallyarguedtoexplainthereluctanceofsomecompaniestoinvest inBigDataispreciselythelackofsuccessstoriesorclearRoIcases.Thatiswhyinvestingindemonstrators(throughlargescalepilotsorlighthouseprojects)andusecasesthatbring the general concept of Big Data to real processes, products and operationalenvironmentsiscrucial.FromtheMonitoringReportofthePPP[4]wecangetagoodoverviewoftheexistingprojects in theportfolioaswellas their focus.Furthermore, interviewsandrepliesfromtheprojectstothisassessmentprocessshowthattheprogramaccountsfor:

• 26 Large Scale experiments (19 involving closed data) and over 150 usecasesandexperiments.

• 4majorsectors(BioEconomy;Transport,MobilityandLogistics;Healthcare;Smart Manufacturing) covered with close to the market large-scaleimplementations,andover10differentsectorscoveredintotal.

ProjectsinthePPPreported151usecasesor/andexperimentsconductedduring2017withacontributionfrom8differentprojects.BDVAmembers,includingthoserunningBDVAlabelledI-Spaces,reportedindependentlyandadditionally(totheBDVPPP projects) 130 data experiments during 2017 most of which have beendeveloped inBDVA I-Spaces. In relation to the amount of datamade available forexperimentation,projectshavereportedatotalof0,0854Exabytes(85,4Petabytes)for2017.



15

Thatisthepictureofwhatwasdevelopedthepreviousyear.However,asecondwaveofprojectsresulted fromthe lastcall,givingbirthtonewinitiatives,manyof themfallingunderthetopicofdataintegrationandthus,ofhighinteresttoDataBench.Oneoftheconclusionswecanextractisthat,eventhoughprojectswillbeafirstsourceofdataforthepurposeoftestingtheDataBenchToolbox,weshouldnotneglectotherinterestingenvironments,alsowithinthePPPframework,whichcouldgiveusaccesstoawideumbrellaofdatasets,sectorsandtypesofexperiments.Thisisespeciallythecase,asmentionedabove,oftheInnovationSpaces,orinshort,I-Spaces.ForabetteroverviewofthecurrentstateofplayofsectorsintheBigDataPPPweadda comprehensive analysismade byBDVA and theBDVeproject inAnnex 1 in thecurrentdocument.Noticethatthisinformationisavailableinthemonitoringreport2017[4],butnopublicversionofsuchdocumentisavailableatthetimeofsubmittingthisdeliverable.Neverthelessweconsiderthisinformationofinterestfortheoverallunderstandingof theportfolioand therefore theenvironmentDataBenchneeds todealwith.

Figure3Portfolioanalysis:mappingwiththeBDVSRIAimplementationmechanisms(source:[4])



16

Figure4VisualrepresentationofthePPPprojectportfolioBenefits(why):

• ForDataBench:environmentforvalidationandtesting;inaddition,potentialcreationofcustomer/userbase

• For the projects: away to identify additional areaswhere benchmarks areneededandredirecteffortstowardsthem,contributetothemodel(research-oriented), get advantage to benchmark the solutions they are producing attechnical level (when BDT are generated) and benchmark vertical pilots intheirperformancewithrespecttobusinessindicatorsthatareaffectedbyBDT

• ForthePPPasawhole:DataBenchcouldbecomeanextremelyinterestingtooltoevaluateandbenchmarkthesolutionsproducedbythePPPandunderstandthe progress with respect to state-of-the-art technologies. As such, theDataBenchToolboxhas thepotential tobecomeavery relevant tool for theprogram (including decisionmakers such as the European Commission) tounderstandtheimpactoftheinvestmentsmadeinBigData.

How(thewaythecollaborationwillbeorganized):DataBenchwill address projects in the PPPportfolio through thewell-establishedgovernancestructureoftheprogram(includedeitherintheGrantAgreementorContractual Agreement of each project); the governance structure governs therelationshipsandcollaborationswithintheportfolio.Suchstructureiscomposedoftwoformalcommittees,thesocallSteeringandTechnicalcommittees.Inthefirstone all project coordinators share knowledge, experience and outcomes of theirprojectsanddefinetogetheractionsinareaslikePPPmonitoring(exerciseknownasthe“monitoringreport”),marketingandcommunications(includingevents)and ingeneraltopicswherewefindcommonalitiesamongprojects(e.g.impactofGDPR).InthecaseoftheTechnicalCommittee,itiscomposedbythetechnicalcoordinatorsof



17

all projects; so far the focus has been on mapping their outcomes to the BDVAreferencemodelasanattempttounderstandtheprogressof thePPPagainst theobjectivesdefinedintheSRIA1[5].Thiscommitteehasalsobeenusedasavehicletogather input for standardizationbodiesatprogram level. The committees includeexpertsfromdifferenttaskforcesofBDVAwhenitisneeded(aclearexampleofthisis precisely the group on standards but also the one on benchmarks). The activeinvolvementofsomeoftheDataBenchpartnersinthemanagementstructure(AtosaschairoftheSC),butalsotheworkingstructureofBDVA(itisofspecialrelevancetheleadershipofTF6on technical aspectsbySINTEF,whereactivitiesonbenchmarkstakeplace)giveDataBenchaprivilegedposition to relate to theprojects.All thesechannelswillbeusedasfollows:

• Steering Committee: to reach out to all project coordinators; provision ofgeneralinformationonevents,achievementsandavailabilityofresults;overallcoordinationandknowledgeexchange

• Technical committee and BDVA TF on benchmarks: they will be used in acoordinatedway.Whilethefirstonewillbeachanneltoaddressallprojectswithtechnicalinfo(inmanycasesinconjunctionwithSC),thesecondonewillbeusedasworkingenvironmenttodevelopcontentsandactionsinadeeperway.

DataBenchwilldeveloppilotswithsomeoftheseprojects.Thiswillrequireacloserrelationship with individual projects and their actors. As such, all projects to beaddressedbyDataBenchwillbedistributedamongDataBenchpartnerstomakesurethat they all have a single point of contact in the project. This will makecommunicationsmoreagileandshouldalsoleadtoatrusted(personal)relationship.

3.2 Innovation Spaces or I-Spaces

Asitwasmentionedbefore,agoodrangeofdata-drivenexperimentsishappeninginthe contextof the I-Spaces. I-Spacesare secured infrastructures forexperimentingwithdata(openandcloseddata)anddifferentanalyticsplatforms.Theydonotonlyprovide technical facilities but also additional services such as legal advice formanagementofcontractualaspectsbetweendataownersandusers.Valueproposition(why):DataBench could help them to assess different platforms and tools and also tounderstand the impact between specific technical architectures/approaches andbusinessbenefitsderivedfromthem.ForDataBench they are an opportunity that isworth exploring, since they couldbecomepotentialbeneficiaries/customersoftheDataBenchframework.WeprovidehereasampleofsomeofthemostrelevantI-Spaces.TheywillbefurtheranalyzedinordertoselectcasesforDataBenchvalidationandwiththeaimofdefiningadditionalexploitation roots.The complete list, labellingprocessand fulldescriptions canbefoundin[2].

1SRIAstandsforStrategicResearchandInnovationAgenda.



18

I-SPACE LOCATION DESCRIPTION PLATFORMS SERVICES

SmartDataInnovation Lab(SDIL)[6]

Karlsruhe

(Germany)

SDIL offers big data researchers access to a

large variety of big data and in-memory

technologies.Industryandsciencecollaborate

closely to find hidden value in big data and

generate smart data. Projects focus on the

strategicresearchareasofIndustry4.0,Energy,

SmartCitiesandpersonalizedMedicine

SAP Hana, Software AG

Terracotta, IBMWatson

Foundation Power 8,

Huawei Fusion Insight,

System:HTCondor

• Infrastructure providing: The infrastructure, including technical

support,isprovidedfree-of-chargebytheSDILoperationpartnersto

anySDILproject.

•Communities:SDILprovidesaccesstoexpertsanddomain-specific

skillswithinDataInnovationCommunitiesfosteringtheexchangeof

project results. They further provide the possibility for open

innovationandbilateralmatchmakingbetweenindustrialpartners

andacademicinstitutions.

• Data curation: The SDIL guarantees a sustainable invest to all

partnersbycuratingindustrialdatasources,bestpractices,andcode

artefacts,thatarecontributedonafairsharebasis.

•DataAnonymization:TheSDILoffersvariousanonymizationtools

to its projects which are applicable to data from research and

industrialsources.

Teralab[7]

Paris

(France)

Teralab isaBigDataplatformtoenhancethe

value of industrial data in partnership with

laboratories or collaborative projects. It has

been operational for more than 3 years and

offersstate-of-the-artinfrastructureandtools.

TeraLab's infrastructure is secure, sovereign

andneutral.Itprovidesthenecessarysecurity

guaranteesforindustrialpartnerssothatthey

can make available, within a defined

framework, theirhigh-valuedata forresearch

orinnovationprojects.

Apache Suite Open

Source on top of in

memoryanalyticsbased

onAtosBulltechnology.

• abilitytoperformexperimentsonmoredata-sets fromdifferent

sectorsacrossEU;

• access for industry from anymember state to SotAexperiment

platforms&tools;

•market-realisticuse-conditionstotest&validatenewtoolconcepts

fromAcademia;

•enablingcross-regionalaccesstoSotAAcademicknowhow;

•sharingbestgovernanceandincubationsupportmethodsbetween

existingi-Spaces;

•wider access to industry data and challenges for education and

trainingofstudents.

Know-Center[8]

Graz

(Austria)

Know-Center is Austria's leading research

center for data-driven business and big data

analytics. It conducts applied and

interdisciplinary research in the field of

computer science in the areas of data-driven

business, big data and cognitive computing.

Main research areas fall under: Knowledge

Discovery, Knowledge Visualization, Social

Computing,UbiquitousPersonalComputing&

BusinessModels, DataManagementandData

Security.

HDFS, MapReduce2,

YARN, Tez, Hive, Pig,

ZooKeeper, Kafka,

Kerberos, Slider,

Zeppelin

Notebook,Jupyter

Notebook, Spark,

Spark2,ApacheSolr.

Theofferincludesconsultations,dataanalysisandtrainings.TheBig

Data Lab offers a simple and direct access to both expertise and

infrastructure. In addition to ApacheHadoop, the Big Data Lab is

equippedwithotherbigdatatechnologiessuchasApacheSparkand

Apache Storm on its computer clusters. Integrated within an

internationalnetworkaroundthetopicofBigDataandDataScience,

the Know-Center provides its partners with access to the latest

trendsandfindingsinthisarea.



19

From the business point of view, main areas

covered by Know-Center are: Industrial Data

Analytics, Data-Driven Markets, Strategic

Intelligence,Data-drivenProcessandDecision

Support,Learning4.0andDigitalLifeScience.

RISE SICSNorth ICE[9]

Luleå

(Sweden)

ICE, the Infrastructure and Cloud research &

test Environment, is a research data center

inaugurated in January 2016. The facility is

open to use primary for European projects,

universities and companies. However,

customersandpartnersfromallovertheworld

are welcome to use ICE for their testing and

experiments.

• HOPS - Hadoop as-a-

service;

•Tensorflow-as-a-

service;

•Streaminganalytics-as-

a-service;

•ApacheSpark;

•ApacheFlink;

• Customized common

development

environment

The range of the offering extends from choosing to just use our

Hadoop application HOPS, all the way to full service with tool

experts,analystsandeventhepossibilitytousethedataownedand

storedbythecenter.ICEoffercoversallpartsofthestack:

•Bigdataandmachinelearning-Computingcapacity,platformsand

toolsforhandlingbigdataandmachinelearning;

•ITandcloud-testingandexperimentenvironmentsforsoftware

development,scalingandinfrastructureoptimization;

•FacilityandITHW-possibilitiesfortestingdisruptiveinnovations

concerningthefacilityandhardwareofadatacenter;

•Utility-measurementsandresearchsecuringasustainablesociety

withefficientdatacentersasapartoftheenergysystem.

EURECAT:Big DataCentre ofExcellence[10]

Barcelona

(Spain)

TheBigDataCentreofExcellenceinBarcelona

isan initiative ledbyEurecatwhichhasbeen

launchedinFebruary2015withthesupportof

the Government of Catalonia, the Barcelona

City Council and Oracle. The Big Data CoE

constructs, evolves, integrates and makes

available to companies differential Big Data-

relatedspecialisedknowledge, tools,datasets

and infrastructures that will allow them to

define,experimentwithandvalidateBigData

modelsandtheirimpactonbusiness,aswellas

define innovative solutions within a

collaborativeframeworkwithkeyagentsfrom

thesector.

DATURAisanOpenStack

based platform

developedbyEurecatto

provision,configureand

deployawholeBigData

stack through an UI

assistant matching

different project

requirements. ATURA

currently allows to

deploy Hadoop

environment clusters

(including HDFS, YARN,

Hive, Pig and Sqoop) as

well as Spark,

Elasticsearch and Kafka

clusters.

TheportfolioofservicesprovidedbyEURECATrangesfromBigData

StrategyandPlanningtoBigDataValueandDiscoveryorBigData

deploymentsupport.

Table1:Descriptionofi-SpacesaspotentialusersofDataBench



20

3.3 Other potential users of DataBench I-SpacesrepresentaveryobviouscaseofpotentialuseroftheDataBenchoutcomes,

butthereareotherinstrumentswithsimilaritiesthatcouldwellbecomecustomersof

theDataBench solution and in a first step, validators/contributors to it. Since our

resourcesare limited theywillbekeptaspartof thedisseminationstrategybut if

some interest is raisedbyourmarketingmaterialwewill address them inamore

focusedway.

In that group we have identified specifically the network of Big Data Centers of

Excellence(CoE)aswellasDigitalInnovationHubs(DIH)withcompetencesonBig

Data,datamininganddatabasemanagement.ThenetworkofCoEcurrentlyaccounts

for55centers.ThecaseofDIHismoredifficulttoquantify,sincethetoolproposedby

JRC-whereDIHcanbechecked-isbasedonentriesfilledinbytheDIHassuch.

Figure5DistributionofBigDataCentersofExcellence[11]

Figure6DIHwithcompetencesinBigData,datamininganddatabasemanagement[12]



21

Finally, as an extension of the networks listed herewewould like tomention the

investmentmadebysomecountriesinBigData,resultinginamyriadofinitiatives

wheredataecosystemshaveemergedatnationallevel.Theseinitiativesoperateas

BDVA, but in their respective countries or regions and in some cases they are

extremely powerful. It is not by chance that some overlaps/synergies take place

between some of the instruments described in this section (CoE/DIH/national

initiatives).Theycanalsobeveryinterestingchannelstodisseminatetheresultsof

DataBenchanddependingonthenatureoftheinitiativetheycouldalsobecomeusers

(thatcouldbethecaseofanI-Space-typeofinitiative)ormultiplytheeffectsofthe

disseminationstrategyoftheproject.

Anoverviewofsuchinitiativesisrepresentedinthefollowingdiagram.Theywillbe

analysedwiththesupportoftheBDVeproject[13]tounderstandtheirpotentialvalue

forDataBench.

Figure7SampleofNationalInitiativesonBigData[13][14]

3.4 Benchmarking communities One of the target communities of DataBench is the one composed by different

benchmarkingeffortsintheareaofBigDataTechnologies,wheremanybenchmarks

forparticularindicatorsoftechnicalnatureexist.AtproposalstageDataBenchalready

identifiedalistofpotentialbenchmarkswewouldliketorelatetothatwereproduce

here.SomeoftheirbenchmarkswillbeintegratedintotheDataBenchToolbox.They

willthereforebedirectcontributorstoDataBench.Inexchangetheywillbenefitfrom

the opportunity to increase their user base. The engagement with them will be

realizedthroughthesettingupoftheBigDataBenchmarkingCommunity(BDBC)establishedbyWP1,whichwillhelptobuildsynergiesbetweenthemaininternational

benchmark communities (such as TPC, SPEC, BigDataBench, Linked Data

Benchmarking Council (LDBC), Hobbit and others) to create a self-sustainable

interactionmodellikelytocontinueaftertheendoftheproject.



22

Figure8ListofbenchmarkinginitiativesanalysedbyDataBench



23

4. DataBench timing: Dissemination Phases DataBenchhasdefinedfourmajorphasesofdisseminationandcommunicationactivitiesthatwill be revisited during the project if needed. Initial inputwas already sketched atproposalstageandhasbeenusedherewithminormodificationsandadaptationsasmainbasisfortheupcomingwork.

4.1 Phase 1 - Create awareness (M1-M12) The initialphasewill focuson creatingawarenessabout theproject, identifying thekeystakeholder communities and buildingworking relationshipswith them (some of theseelementshavebeendepictedinprevioussections).Keytothisobjectivewillbeontheonehand theBigDataBenchmarkingCommunity(BDBC) thatwill allowDataBenchtoworkhand-in-handwith themost relevant benchmarking initiatives, but also the relationshipwithusecasesandpilots,especiallythroughtheconnectionwiththePPPprojectportfolio.Froma communicationpointof view the strategywill focusonensuring thatasmanypeople and organisations as possible that are stakeholders in the Big Databenchmarkingandbusinessanalysisarenaknowabouttheproject,itsobjectivesaswellastheexpectedbenchmarksandmetricsthatwillbedelivered.Thiswillincludepresentationsandvisibilityat someBigData conferencesandworkshopsaswell as theelaborationofthefirstdisseminationkitcomprisingatleastaprojectfiche,handout/flyer,roll-upandaposter,besidesaDataBenchpresentationtemplate.Thismaterialwillbeusedinternallytoensurethattheprojectteamcommunicatesthesamecoherentmessage,butalsoexternallytoassessthepotentialacceptanceoftheconcept.Thiswillbeassessedandinputwillbelateronusedtorefinemessages.Thewebstrategywillalsobelaunchedinthisinitialphase,openingourmain“window”totheworldthroughthewebsiteandfirststepsinsocialnetworks.

4.2 Phase 2 - Increase the potential impact (M13-M24) During the second year, the project will deliver the preliminary benchmarks, the firstmetricsandtheAlphaversionoftheToolbox.Thedisseminationstrategywillthenfocusoncreatinganappealingand“marketable”versionoftheresults,andtocommunicateandsharethemwiththestakeholdercommunitiestargetedinPhase1,increasingthewillingness to participate in the test and validation of these methods. The project willorganise webinars and participate in community events to support the collaborativeworking relationships establishedwith the projects and other instruments thatmay beusefultothispurpose(asidentifiedinsection3ofthisdocument).Usageofchannelslikenewsletters or blogs will be maximized. This phase is expected to leverage the projectwebsiteandsocialmediatoensurealsothewiderdisseminationofresultsandtoattracttheattentionofpotentialusersoftheToolboxandtheproposedbenchmarks.

4.3 Phase 3 - Maximise Results (M25-M34) Duringthenexttenmonths,astheprojectdevelopsthefinalversionofthebenchmarksandtoolbox,themainobjectivewillbetomaximisethepotentialimpactsbycontinuingthecommunicationandnetworkingactivities,asdepicted inPhase2,butwitha focusonbuildingconsensusandrecognition.Beyonduseofwebinars,participationinworkshopsandconferences,wewillalsobegintopublishtechnicalpapersdetailingtheBenchmarksand the Toolbox, which will be presented in at least three international Big Data



conferences. In the final year, we will organize specific community events to presentcompleteDataBenchfindingsandtorundemostoexhibittheeffectivenessoftheToolbox.TheprimarygoalistoincreasetheimpactthroughtheexternalbenchmarkinginitiativesandtoattractpotentialusersorclientsincludingBigDataindustryandpolicymakers.Thiswillconstitutethe“DataBenchproductroadshow”.

4.4 Phase 4 - Valorisation (M34-M36+M>36) Fromthenearend(finaltwomonths)oftheprojectandaftertheformalendoftheprojectduration,demonstrationsforspecificaudiences(researchcommunity,industry,policymakers) will be carried out. In the same period, the contents and findings related toHorizontal Benchmarks, Benchmarks of European and industrial significance, andDataBenchHandbookassociatingtheDataBenchFrameworkandtheDataBenchToolboxdescribing metrics implementation and benchmarks will be widely published innational/international journals and in on-line media. The primary goal will be todemonstrate to internal and external potential adopters, existing Big Data researchcommunitiesandpolicymakers the relevanceofbenchmarkingapproaches forBigDataTechnologies.



5. Implementation of the Strategy: tools and channels This section describes the “how-to” part of the strategy, essentially channels, tools andmechanisms that will be used by DataBench to create awareness and disseminate thedifferentmessagestothecommunitiesofinterestthathavebeenpreviouslydescribed.

5.1 DataBench branding The first step consists in the definition and agreement on a branding that reflects theidentityofDataBench.Thiswasdoneatthebeginningoftheprojectwiththesupportofprofessionaldesignersresultinginseveralcreativeelements,includingthelogo,whosefinalversionis:

Figure9DataBenchlogo

Detailsaboutthemeaningoffont,coloursandotherelements,aswellastheprocessbehindtheartisticcreationcanbefoundinD6.2.

5.2 Marketing Material Oncetheidentityiscreated,alltheothermaterialsshouldfollowthesamegraphicalimage.DataBenchhasprioritizedtheelaborationofthefollowingmaterialforthefirstphaseoftheproject:projectfiche,handoutandroll-up.Thisiscomplementedbyaseriesoftemplatesthatgofromdeliverablestopresentationsandpressreleases.AllofthemcanbecheckedinD6.2.Thisinitialmarketingtoolkitshouldbeenoughatthisstageoftheprojectwheremaincontentisbeingcreated.Forthesecondpartoftheyear,andlookingatsomeimportantandbigeventsoftheresearchcommunitywhereDataBenchwillbepresent(EBDVF,ICTevent),wemaycreateanadditionalposterfocusedonspecifictechnicalcontentsandavideo.Wearealsointheprocessofselectingacatalogueofgoodiesthatshouldservethepurposeofattractingvisitorstoourbooth/exhibitionareaineventswherealotofcompetitorsfighttogettheattentionofpeople.

5.3 Strategy on the use of web channels GeneraloverviewDataBenchwilladoptamulti-channelonlinestrategyinordertoimplementdisseminationinitiativesandengagewiththestakeholders,maximisingitsimpactandvisibility.Thewebchannelsincludetheprojectwebsite(includingthecommunityspaceandToolBoxtestingarea), the socialmedia channelsand thenewsletter.Abriefdescriptionof all of them isprovidedhere,althoughweredirectinterestedreaderstoD6.2,wheretheycanfindmoreinformation.



ProjectwebsiteTheprojectwebsite,availableat theURLwww.databench.eu, isbasedontheopensourcecontentmanagementsystem(CMS)byWordPress that allows users with an account to create webcontentviaatext-basedinterfacethroughawebbrowserwithoutanyHTMLeditinginvolved.Thewebsitepresentsthevisionandthemostimportantinformationabouttheproject;asitwassaidbefore, it is considered ourmainwindow to theworld, and as

such,italsoprovidesaccesstoallpublicmaterials.Findbelowaglimpseonmainsectionsandinformationthatcanbefoundinthecurrentversion:

• Home page: designed with an attractive and eye-catching layout to engage thevisitors,andinlinewiththevisualidentityoftheproject,itgivesanoverviewofthefocusof theproject, aswell asquickaccess to the latest contentuploadedon thewebsiteandpostspublishedonthesocialmediachannels.

• Theproject:thissectionprovidesashortdescriptionoftheprojectscope,itsmainobjectivesandexpectedoutcomes.Italsopresentsprojectpartners.

• Resources: this section is a repository of the public materials produced by theproject, available for visitors’ download. The resources available include publicdeliverables,scientificpublications,presentations,marketingmaterialsandwebinarrecordings.

• Newsandevents: thissection features theupdatesabout thedevelopmentof theprojectanddisseminationactivities,intheformatofjournalisticnewsandblogposts.Italsogivesinformationabouttheparticipationandcontributionstoexternaleventsand/orevents/workshop/sessionsorganizedbytheproject.Acalendarofrelevantevents,includingwebinars,isalsoavailableinthissection.

• Communityspaceandtestingarea:oneofthekeyoutputsoftheprojectwillbetheDataBenchToolbox,asmentionedseveraltimesalongthisdocument.Itisthereforeimportant to create a place or area on the website fully devoted to it and, inparticular,tothepotentialtestingofthetool.Thewebsitewillfeaturearestrictedarea,availableuponregistrationandapprovaloftheprojectteam,whichwillgiveaccesstoacollaborativespaceandtotheDataBenchToolboxfortestingpurposes.The DataBench Handbook providing guidelines on when and how to use theDataBenchToolboxwillbemadeavailableintherestrictedarea.

Theprojectteamwillperiodicallyupdatethecontenttomakesurethatthewebsiteprovidesup-to-dateinformation.

SocialmediaDataBenchwillestablishastrongsocialmediapresencethroughthe following platforms to boost the dissemination andcommunicationactivities.Socialmedia channelswillbeused toconnectwiththecommunityofstakeholdersandtodrivetraffictotheprojectwebsitethroughlinkstothecontentuploaded.

• Twitter: thisplatformhasbeen identifiedasoneof themostpowerfulmeanstoconnectwiththeresearchcommunityaswell

aswithotheraudiencegroups.Itwillalsobeusedtoattractattentiontotheactivities



oftheproject,todrivetraffictothewebsiteandtocross-publishanyrelevantthird-partycontent.Theaccounthandleis@DataBench_eu.

• Facebook: the project team is aware that Facebook is less used by our targetaudience,howevertheDataBenchpagewillbe leveragedtoreachout toboththegeneral public as well as to the targeted stakeholder groups that are also onFacebook.Theaccounthandleis@DataBenchEU.

• LinkedIn: an open group (DataBench Project) has been created to facilitate thediscussionwiththedatacommunityinaforum-formatapproach.

• YouTube:adedicatedchannel(DataBenchProject)hasbeencreatedtoshareanyvideoproducedbytheproject,includingtherecordedvideosofthewebinars.

• SlideShare: the project account (DataBench) will be used to periodically shareDataBenchpublicpresentations.

All social media channels reflect the visual identity of the project. Each post will beaccompaniedbyvisualcontentandwillencourageengagement.

Withtheaimofgeneratingtraffic,thelinktotheDataBenchwebsiteisdisplayedonallsocialmediaprofiles.

NewsletterAt the time of submitting this document the projectacknowledges that newsletters are important tools to informpeopleinterestedinthetopiconaperiodicalbasis.Theyenabletheinclusionofmoredetailedinformationabouttheproject,thusengaging the community in amore solidway.Main objectivespursuedbyDataBenchwouldbe:

• RaiseawarenessaboutDataBenchanditsactivities,inparticularinthefirstphaseoftheproject;

• Giveregularupdatesontheprojectactivities;• Drivetraffictotheprojectwebsitewherethecompletetextofthenewsletterwillbe

available;• Increasecommunityengagement.

Nevertheless,DataBenchisstillanalysingthesuitabilityofintegratingthecontentintotheBDVAoneorhavingitsownnewsletter.Thefirstoptionseemsmorepromising,sincewewouldtargetimmediatelymorethan1000recipients(closeto2KbeforeGDPR).Atthetimeof submitting this document, DataBench has already contributed to the BDVA/PPPnewsletterwithspecificcontent.

5.4 BDVA tools and engagement with the PPP community EventhoughwehavealreadytackledthecollaborationwithprojectsofthePPPportfolioinaprevioussectionofthisdocument,itisworthgoingdeeperintothepossibilitiesofferedbybecoming part of a program. In this particular case,DataBenchwill capitalize the tightrelationshipswith the BDVAworking structures and the opportunitiesmanaged by theBDVeproject through the communications committee (not a formal committee though),whichworksinalignmentwithBDVA.Mainchannelsandtoolsofinterestidentifiedatthebeginningoftheprojectandthatwillbeworkedoutare:

• BDVAwebsite,PPPportal,newslettersandsocialnetworks:BDVeisresponsibleforthecommunicationactivitiesatprogramlevel,andaspartofthatresponsibility



it runs the Big Data Value Portal, a placewhere you can find descriptions of allprojects,amongotherthings.DataBench informationhasbeen includedaspartofthis ecosystem and the project contributes to the different communicationcampaigns launched in that framework.Oneof themostsuccessfulones is called“projectoftheweek”,whereaprojectisselected,andmoredetailedinformationisprovidedtothePPPcommunitysothatpeoplegettoknowitindepth.Thisportal,which also provides access to different tools (as you will see below) iscomplementary to the BDVA website, which is used as additional channel forpromotion.BDVemanagessomeofthewebtoolsofBDVA/PPPandinparticularthenewsletterandsocialnetworkchannels.Alotofattentionispaidtotheprojectsoftheprogram,andtheyareheavilyusedtohighlighttheirevolution,news,outcomes,etc.Thisisaclearadvantagetoanyproject,sincethesetoolsopenupthedoortothousandsofpeoplethatotherwisewouldbedifficulttoreachoutbyasingleproject.Theuseofprogram-leveltoolsisthereforeencouraged.Assuch,DataBenchhasalreadysentoutinformationandusedthesetoolsinthelastmonths.

Figure10DataBench@PPPportal[13]

(DetailofthesectiondevotedtoPPPprojects,whichrunsanalgorithmtoensurethatprojectsareshowninaleatorypositions)

• The Big Data landscape: this tool will be accessible from the PPP portal andrepresents major actors of the data economy in Europe on a map. It builds onpreviousefforts,suchasthedatalandscape.eudevelopedbyIDCandOpenEvidence.



Someoftheadvantagesofthelandscapearethatitallowsuserstomakeintelligentsearchbyusingmultiplefiltersanditalsofeaturesdifferentlayersofinformation,including stakeholders that are not single organizations but initiatives that couldenable other actors to get engaged in the Data economy. Examples of enablersrepresented on the map are the network of Big Data Centers of Excellence, amultiplicityofnationalinitiativesandprogramsfocusedonBigDataandtheexistingnetworkofdata-drivenexperimentationenvironmentsknownasi-Spaces.Themapwillalsoshowthedistributionofpilotsimplementedintheprogram.Assuch,itwillbeausefultoolforDataBencheventhoughmoresuitedforconnectionsthanforpuredisseminationpurposes.

Figure11TheBigDataLandscape

• TheInnovationMarketplace.ThistoolprovidesvisitorstothePPPportalwithacatalogueofBigDatasolutionsthataresearchablebyusingdifferentcriteria,suchasTRL, domain/sector, typology of solution, etc. It will include all the outcomesproducedbyPPPprojects.TheDataBenchToolboxcouldobviouslybeoneofthoseeventhoughthewayitwillbeexposedandmarketedwillbediscussedlateronintheproject,asitcouldhaveahighimpactonexploitation.

Figure12ThePPPInnovationMarketplace



• ThePPPWebinarSeries:fromNovember2018,theBDVeprojectwilllaunchaverypromisingtool tospreadtheknowledgegenerated inthePPP.Webinarswillgivevisibility to the project results, but also to topics of general interest to the PPP.Speakerswillbe fromPPPprojects,BDVAbutalsoexternalexperts.TheWebinarSerieswillhavetwotracks:atechnicaloneandanindustrialtrack.Someofthepre-selectedtopicsinclude“TipstoapplyGDPR”(i.e.howdoesGDPRaffectaprojectora sector, practical examples, implications, etc.), “Data sharing architectures andsolutions”,“Datavalueandmonetization”orwideinformationaboutexistingdatasources inEurope, as it is the caseofCopernicus/DIASplatforms/EuropeanDataPortal.ItisourpleasuretoannouncethattheTechnicalTrackwillbekickedoffwithDataBench.

Figure13ThePPPWebinarsSeriestobelaunchedinAutumn2018withDataBench

• Events:aspartoftheCommunicationsCommittee,DataBenchwilltakeplaceinthemeetingsandactionsleadingtoaglobalpresenceofthePPPinevents.Thiswillbeparticularlyimportanttomanagethepresenceoftheprojectsinbigeventswhereitisdifficulttogetavisibleplaceeitherintheprogramorintheexhibitionareaifyougoasasingleproject.CriticalmasswillbeusedtogetbetterexposureinPPPflagshipevents,suchastheEBDVForPPPMeet-ups,EC-relatedeventssuchastheICTeventorindustrial-likeevents,asitcouldbethecaseofCEBIT,whereeconomicreasonsprevent a single project from being represented there. Some of these events arementionedlateroninsection5.5.

5.5 Events DataBenchhasalreadyidentifiedalistofeventsofpotentialinteresttotheproject.Theywillbefurtheranalysedduringthecourseoftheprojectanddecisionswillbemadebasedontheavailabilityofresults(dependentontheimplementationphaseoftheproject),budgetconstraintsandopportunitiesthatcouldbematerialized.



Duetothefactthatthisplanissubmittedlaterthaninitiallyexpected,wecantakeadvantageof theavailabilityofadditional informationabouteventsin the firstperiod.AnumberofopportunitieshavealreadybeenexploitedbyDataBench.TheywillbereportedindetailinD6.3DisseminationandLiaisonReport (M18).However,weanticipatehere someof theactionsalreadycarriedoutbytheproject.

DataBenchinH12018InternationalConferenceonPerformanceEngineering(ICPE)[15]

• When:9-13April• Where:Berlin(Germany)• Shortdescription:ICPEintegratestheoryandpracticeinthefieldofperformance

engineering by providing a forum for sharing ideas and experiences betweenindustryandacademia.

• Role of DataBench: Within the conference program we can find multipleworkshops, being one of them the 4th International Workshop on PerformanceAnalysisofBigdataSystems(PABS)[16].ItispreciselyinthisonewhereDataBenchwasrepresented.

BigDataPPPMeet-upSofia[17]

• When:14-16May2018• Where:Sofia(Bulgaria)• Shortdescription:FirsteditionofaPPPeventthatcomestostay.Itisorganizedby

theBDVeprojectandBDVAwiththeaimofbringingtogetherthemaincommunitiesimplementing the Big Data PPP: projects and BDVA membership/Task Forces(content-drivenworkinggroupsoftheassociation).Theeventwascomposedbytwoworkingdays(withagoodnumberofworkshopsrunninginparallel)andanOpenDayintendedtogivevisibilitytothedataecosysteminBulgariathatwassupportedbytheCommissionaireGabriel.

• RoleofDataBench:DataBenchwastheorganizerofaworkshoptogetherwiththebenchmarkingTFofBDVA.Itwasthefirstopportunityfortheprojecttorelatetotheotherprojectsoftheprograminaninteractivemode.

Figure14BigDataPPPMeet-upSofia,theplacewhereDataBenchkickedoffcollaborationwiththePPPprojects



BigDataInnovationConference[18]

• When:7-8June• Where:Frankfurt(Germany)• Short description:The conference is intended tohelp companies understand &

utilizedata-drivenstrategiesanddiscoverwhatdisciplineswillchangebecauseoftheadventofdata.

• RoleofDataBench:Presentationoftheprojectattheconference.

KDD2018(24THACMSIGKDDCONFERENCEONKNOWLEDGEDISCOVERYANDDATAMINING)[19]

• When:19-23August• Where:London(UK)• Shortdescription:Withseveralthousandsofvisitors,KDDisoneofthemostwell-

known international conferences of the data community. The 2018 edition hascountedonkeynotespeakers,invitedtalks,27workshops,8hands-ontutorialsand29conventionaltutorials.

• Role of DataBench:DataBench was one of the projects selected for the ProjectShowcaseTrack,whichofferedafulldayagendafocusedexclusivelyoninnovativeprojects on Machine Learning and Data Analytics to show the state-of-the-art inresearch and applications in this field. The selected paper “DataBench: EvidenceBased Big Data Benchmarking to Improve Business Performance” together withadditional information will be made available for download on the DataBenchwebsite.

DataBenchPrioritiesinH22018-H12019(short-termplan)ItAIS2018:15thconferenceoftheItalianChapterofAIS(AssociationforInformationSystems)[20]

• When:12-13October• Where:Pavia(Italy)• Shortdescription:the2018editionofItAISwillrunundertheslogan“Livinginthe

digitalecosystem:technologies,organizationsandhumanagency”.• RoleofDataBench:Presentationofacceptedpaper.

BarbaraPernici,ChiaraFrancalanci,AngelaGeronazzo,LuciaPolidori,StefanoRay,Leonardo Riva, Arne Jørgen Barre and Todor Ivanov, “Big Data key performanceindicators”,acceptedforpresentationattheXVeditionoftheitAISconference,PaviaonOct12th-13th2018.

ThepaperwillbemadeavailablefordownloadontheDataBenchwebsite.EuropeanBigDataValueForum(EBDVF2018)[21]

• When:12-14November• Where:Vienna(Austria)• Short description: EBDVF is a key European event for industry professionals,

businessdevelopers,researchers,andpolicymakerstodiscussthechallengesand



opportunitiesoftheEuropeandataeconomyanddata-driveninnovationinEurope.Thisyear,theeventwillbepartoftheEUPresidencyagendaandwillcountonthepresenceofCommissionaireGabriel,amongotherrelevantspeakers.TheprogramwilldevotetwodaystokeynotesandparallelsessionswheremanyprojectsofthePPP will be represented. A third day will enable projects to interact inworkingworkshopsresultedfromanOpenCallforcontributions.

• RoleofDataBench:DataBenchisoneoftheorganizersofthesession1.8onBigDataBenchmarking2. Two speakers from the project are already confirmed there. AnadditionalworkshoptocollaboratewithPPPprojectswasalsoproposed,butduetotheacceptanceoftheBenchmarkingsessiononthemainprogram,theworkshopwasdiscarded.DataBenchiswaitingforinstructionsregardingtheexhibitionareaincasewecangetvisibilitythroughthePPPbooth.

ICTEvent2018(ImagineDigital-ConnectEurope)[22]

• When:4-6December• Where:Vienna• Shortdescription:ICT2018willfocusontheEuropeanUnion’sprioritiesinthe

digitaltransformationofsocietyandindustry.ItwillpresentanopportunityforthepeopleinvolvedinthistransformationtosharetheirexperienceandvisionofEuropeinthedigitalage. AsageneralICTevent,itwillfeaturemanysessionsandworkshops focusedonmanydifferent topics, someofwhichwill for sureaddressdata-relatedtopics.

• RoleofDataBench:DataBenchsubmittedaproposalforanetworkingsessionthathasnotbeensuccessful.NeverthelesswewilltakeadvantageofthesessionsproposedbytheBigDataPPPtogetherwithotherinitiativesandplatformswiththeaimof exposingDataBench.NegotiationsareongoingwithBDVe/BDVA topromoteaspeaker in networkingsession(1)of the followinglistofacceptedproposals:

1. "Impact of Data-Driven AI in business sectors” in collaboration witheuRobotics/SPARC PPP, which will convene the 4 BDV PPP lighthouseprojects;

2. "Data democratisation: empowering the citizens in the digitaltransformation” in collaborationwithMyData community, Open and AgileSmartCitiesandtheLivingLabs(ENoLL);

3."Personalandmachine-generateddata:impactonprivacyandsecurity”incollaborationwithECSO;

4.“DataScienceSkillsforSociety:FromBigDatatoAI,roboticsandbeyond”doneincollaborationwitheuRobotics/SPARCPPP;

5. “Common priorities of HPC, Big Data and HiPEAC for post-H2020 era”,resultofthecollaborationbetweenETP4HPC,BDVAandHiPEAC;

2Seehttps://www.european-big-data-value-forum.eu/program/



6."Wouldyoubuyasecond-handcarfromthisalgorithm?”(EvolutionoftheFateofAI,ellaboratedbyBDVAincollaborationwithRoboticsPPP).

A booth for the PPP has also been accepted. Itwill be themeeting place for PPPprojectsattheexhibitionarea.Thatiswhyithasbeencalled“TheBIGDATAVALUEPublicPrivatePartnershipvillage”.DataBenchwillbepartofthevillage.

OveralllistofeventsofinteresttoDataBenchobjectivesThe following table filters some of the events that have been selected as potentialopportunitiesforDataBench.Theywillberevisedonacontinuousbasisinordertoconfirmwhichoneswillbeattendedbytheproject,actionsneededtocarryout theworkandtoupdatethelistwithemergingopportunitiesthatwemayhavenotconsideredsofar.Thelistshouldthereforenotbetakenasacompleteorstableone.Eventspointedoutaboveandalreadyconfirmedhavenotbeenincluded.

Event Whenandwhere Furtherinformation

CEN-CENELECStakeholderWorkshop -TrustworthyArtificialIntelligence

Brussels(Belgium);18September

Standardizationfortake-upofAItechnologies;

DataforAI Brussels(Belgium);19September

The focus of this workshop, jointly organized between the EC,BDVA and the euRobotics platform is on identifying the mainchallenges to develop the European Data Space as essentialcomponentforAI;

12thEuropeanConference onSoftwareArchitecture(ECSA)

Madrid (Spain); 10October

Conference that brings together major experts on softwarearchitectureresearchandpractice.Oneof theworkshops,whichwill count on participation of BDVA is of special interest to us:“Workshop on Software Architecture Challenges in Big Data -SACBD@ECSA2018”;

High TechSummit

Copenhagen(Denmark); 10October

ADTU3-poweredeventthatisthelargestresearch-basedmeetingplaceinDenmarkwithin the fieldsofdigitalization, roboticsandartificialintelligence;

6th GlobalSummit onArtificialIntelligenceand NeuralNetwork

Helsinki (Finland);15October

Scientific conference addressing knowledge and sharing of newideas amongst the professionals, industrialists, researchers, andstudentsfromresearchareaofArtificialIntelligence;

IoT SolutionsWorldCongress

Barcelona (Spain);16October

Major meeting point for people interested in IoT for DigitalTransformationorIndustrialIoT;

3TechnicalUniversityofDenmark



ODBASE 2018- The 17thInternationalConference onOntologies,DataBases,andApplicationsofSemantics

La Valetta (Malta);16October

Forumfocusedontheuseofontologies,rulesanddatasemanticsinnovelapplications;

BigSurv 2018:Exploring newstatisticalfrontiersattheintersection ofsurvey scienceandbigdata

Barcelona (Spain);25-27October

Thisevent hosted by theEuropeanSurveyResearchAssociation(ESRA) addresses how researchers produce, analyze, and usestatistics;

Table2:ListofpotentialeventsforDataBench



6. KPIs and Monitoring Framework At proposal stage DataBench described some of the targets for marketing andcommunications. Theywill be used as starting point tomonitor the performance of theproject. Theywillbe revisitedduring theproject andwill consider thedifferentphasesexplainedinChapter4,sinceexpectationsinnumberswillnotnecessarilygrowinalinearway.D6.3,as thedeliverablethatwillreport theactivitiesdevelopedbytheproject,willaddressthisaspectinamoredetailedway.

Activity Indicator TargetKPI

Technicalpapers Numberofpapersandnumberofconferenceswherepapersarepresented

Atleast5paperspresentedinatleast3internationalBigDataconferences

Projectwebsite N.ofuniquevisitorstothewebsite(averageperyear)

Min.2000

Socialmedia-Twitter

N.offollowers

Newfollowersperyear

Y1300

+100

Socialmedia-Facebook

N.offollowers

Newfollowersperyear

Y150

+100

Social media -LinkedIn

N.offollowers

Newfollowersperyear

Y1100

+100

Socialmedia-YouTube

N.ofvideospublished

N.ofviews

Min.4

100viewspervideo

Socialmedia-SlideShare

N.ofoverallviews 200

Newsletter N.ofsubscribersperyear

Numberofnewsletters

100

3peryear

Webinars N.ofwebinars

N.ofparticipantsperwebinar

Min.4

10

Table3:DisseminationandcommunicationKPIs



7. Conclusions DataBenchhasthepotentialtobecomeakeyelementoftheBigDataValuePPP,sinceitwillcontributetomeasuretherealimpactoftheinvestmentsmadebyindustryandtheECinthispartnership.ItwillsupportthebenchmarkingofBDT,thereforehelpingtounderstandtheprogresswithrespecttothestateoftheart,butwillalsorelatetechnicalperformanceindicatorswithbusinessindicators,establishingabridgethathasneverbeenbuiltbefore.

The challenges are however important. Among them, DataBench will require tightcollaboration with a number of players and communities. Working hand-in-hand withbenchmarkingcommunitiesaswellaswithpilotsandusecases(ideallycomingfromPPPprojects) iskey.Theywillbecontributorsbutalsovalidatorsof thesocalledDataBenchToolBox,mainproducttobeproducedbytheproject.

Along the three years duration of DataBench we will have to gain the credibility andrecognitionofthosecommunities,since-inmostcases-theywillbethefirstadopters/usersoftheDataBenchoutcomes.

WP6 will support DataBench in creating the connections and engaging with thosecommunities,willdefineandwillimplementadisseminationandcommunicationstrategyandwill pave the path towards the self-sustainability of the results beyond the projectduration. This deliverable describes the initial steps in that endeavour by sharing theDissemination and Liaison Plan. The document identifies specifically: (1) targetcommunities/audience, (2)Phasesof thedisseminationstrategy, (3) tools, channelsandmechanismsthatwillbeusedand(4)KPIs.

Among the target communities it is worth mentioning the PPP project portfolio,benchmarking communities, I-Spaces (data-driven experimentation environments) andnetworksofresearchorinnovationstructuresthatcouldwellbecomepotentialusersoftheDataBenchToolBox(e.g.BigDataCoE,DIH).

Due to the submission of this document later than initially planned, we have takenadvantage not only to identify actions but also to report some of the activities alreadycarried out by the project in different areas: web strategy, publications, or visibility atevents.



8. References

[1]D3.1DataBenchArchitecture

[2]http://www.bdva.eu/

[3]Contractual arrangement: Setting-upaPublic-PrivatePartnership in theareaofDatabetween the Big Data Value Association and the European Commission (Brussels, 13October2014)

[4]MonitoringreportBigDataPPP2017

[5] Strategic Research and Innovation Agenda BDVA v4;http://bdva.eu/sites/default/files/BDVA_SRIA_v4_Ed1.1.pdf

[6]https://www.sdil.de/en/

[7]https://www.teralab-datascience.fr/en/home/

[8]http://www.know-center.tugraz.at/

[9]https://www.sics.se/

[10]https://eurecat.org/es/eurecat/centros-de-excelencia/big-data-coe-barcelona/

[11] Big Data Labs Europe: http://magazin.know-center.tugraz.at/i-know-2016/program/big-data-labs-europe/

[12]http://s3platform.jrc.ec.europa.eu/digital-innovation-hubs-tool

[13]http://www.big-data-value.eu/

[14]BDVeproject.D3.11ReportonBigdataNationalandRegionalOutreach

[15]https://icpe2018.spec.org/conference-program.html

[16]https://web.rniapps.net/pabs

[17]http://www.big-data-value.eu/big-data-value-meet-up-sofia/

[18]https://pgsolx.com/IT-Tech/BDIC/

[19]http://www.kdd.org/kdd2018/

[20]http://www.itais.org/conference/2018/

[21]http://www.european-big-data-value-forum.eu/

[22] https://ec.europa.eu/digital-single-market/en/events/ict-2018-imagine-digital-connect-europe

[23]BDVeproject.D3.3.UserEcosystemCharacterization



9. Annex I. State of play of sectors in the Big Data PPP [4] [23] Aspartof theworkdevelopedfor themonitoringreport,BDVA,supportedbyBDVe,hasgathereddatatounderstandthecurrentsituationofactivitiesassociatedtothedeploymentandevaluationofBigDatainconcreteindustrialsectors.Inputcomesessentiallyfromtwosources:thelargescalepilots(alsoknownaslighthouseprojects)oftheprogramandthesectorialworkinggroupsthatareoperatinginBDVA(underwhatitisknownTaskForce7onApplications).Theinformationprovidedherereflectsthecurrentstateofplayandhasbeen extracted directly from [4] with just minor changes. You can find thisinformationdirectlyin[23],whichisofferedhereforcontextawareness.

Transport,mobilityandlogisticsRelevanceoftheSectorBig Data will have profound economic and societal impact on mobility and logistics.Mobilityandlogisticsisoneofthemost-usedindustriesintheworld-withamarketsizeofabout1305B€contributingtoapproximately15%ofGDPandtoemploymentofaround 11.2 million persons in EU-28, i.e., some 5.0 % of the total workforce. TheimprovementsinoperationalefficiencyempoweredbyBigDataareexpectedtoleadto500billionUSDinvalueworldwideintheformoftimeandfuelsavings,aswellassavingsof380megatons CO2 in mobility and logistics. With freight transport activities projected toincrease,with respect to 2005, by 40% in 2030 and by 80% in 2050, transforming thecurrentmobilityandlogisticsprocessestobecomesignificantlymoreefficient,willhaveaprofoundimpact.ThelogisticssectorisideallyplacedtobenefitfromBigDatatechnologies,asitalreadymanagesmassiveflowofgoodsandatthesametimecreatesvastdatasets-a10%efficiencyimprovementwillleadtoEUcostsavingsof100B€.There are huge untapped opportunities for improving operational efficiency, deliveringimprovedcustomerexperience,andcreatingnewbusinessprocessesandbusinessmodels,yet only 19% of EU mobility and logistics companies currently employ Big Datasolutionsaspartofvaluecreationandbusinessprocesses.ThemobilityandlogisticssectorisreadytoexploitBigDatasolutionstosignificantlyincreasetheEUmarketinthemobilityandlogisticssector.Ascitedabove,mobilityandlogisticsisoneofthemost-usedindustriesintheworld,contributing significantly toGDP and employment.However,with total goods transportactivitiesinEU-28estimatedtohaveamountedto3,768billiontonne-kilometres4and6,391billionperson-kilometres(onaveragearound12,652kmperperson),mobilityandlogisticsisalsoakeycontributortoCO2emissionswithtotalgreenhousegasemissionof4,824 megatonnes CO2. Therefore, any improvement in efficiency will have profoundimpactonsustainability.Thismeansthemobilityandlogisticssectorwillhugelybenefitfromtheintroduction,adoptionandlarge-scaletake-upofBigDatasolutions.Asmentionedabove, BigData solutions are only used by a small fraction of EUmobility and logisticscompanies, thuspresentingahugeopportunity for increasing themarket share ofBigDatasolutions.Inaddition,thenatureofthemajorityofmobilityandlogisticsdatadoesnotimplymajor legal barriers to overcome before adaption for data-driven innovation andimmediatetakeupofBigDataValuePPPoutcomes.Asaconsequence,themobilityand

4Atonne-kilometreistheproductoftransportedmassintonnesandthedistanceinkm.



logistics sector - compared with other sectors (such as healthcare, finance, media, ortelecommunications) - is optimally positioned and ready to deliver value fromBigDatasolutions.NatureofDataAssetsintheSectorThedatacharacteristicsexpectedforthetransport,mobilityandlogisticssectorhavebeenpresented in a recent study, inwhich 69%of corporate executives named greater datavariety as the most important factor, followed by volume (25%), with velocity (6%)trailing- indicatingthat thebigopportunity lies in integratingmoresourcesofdata,notbiggeramounts.Comparedwithothersectors(suchashealth,finance,ortelecommunications),themobilityandlogisticssectorfaceslesslegalbarriersforwhatconcernstheuseofdatatodelivervalue; e.g., due to the fact that the many of the available data sources do not includepersonal-identifiabledata.Thismeans themobilityand logisticssector - comparedwithothersectors-isverywellpositionedandreadytotakeupBigDatasolutions.Nevertheless,the European Data Protection Supervisor has identified general concerns related topersonaldatainthemobilityandlogisticsdomain,whichshouldbeaddressed.Reasonablesafeguardsshouldbeappliedtoprotectthepersonalinformationfromunauthorisedaccess,loss,misuse,modificationanddisclosure.Asanexampleoftheindicativevolume,velocityandvarietyofdataassetsintheapplicationdomainwearereferringtoaveragenumberscomputedfromthe13pilotsoftheICT-15BigDataPPPLighthouseprojectTransformingTransport:60,000GB(volume),24.5GB/day(velocity),22differentdatasources.ImpactMobility and logistics stakeholders are already adopting technology and massivelystoringdata.Logisticsprovidersalreadynowmanageamassiveflowofgoodsandatthesametimecreatevastdatasets.Thevolume,velocityandvarietyofdatageneratedtodayisunprecedented - itwill fundamentally transform the sector.Today theseadditionaldatasourcesaugmenttraditionalsourcesoftransportdatacollection,whileinthefuturetheywill likely replace them. From the transport point of view, vehicles are massivelyincorporatingtelematicssystemseitherasOEM(OpenOn-start,BMWconnecteddrive,MercedesEmbrace)orasexistingplug-and-playandretrofittingsolutionsavailableinthemarket(Openmatics,Geotab,MobileDevices,TomTom,etc.).Improvementsintransportprocesseswillhavesignificantsocietalimpact,inparticularintermsofCO2emissions.Itisestimatedthata10%efficiencyimprovementwillleadtoEUcostsavingsof100B€andtimeandfuelsavingsof90megatonsCO2withintheEU.In addition Big Data applied to transport will produce a safer mobility in the EU,contributingtoreducedeathsinthetransportsector(26.000peopledieeveryyearintheEUintrafficaccidents).Themobilityandlogisticssectorfaceslesslegalbarriersforwhatconcernstheuseofdatato deliver value; e.g., due to the fact that themany of the available data sources do notincludepersonal-identifiabledata. In transportand logisticsector,bothtechnologiesanddatasetsaremuchmoreharmonisedandstandardisedcontributingtomuchfasterpan-EUimplementationandadoptionofbigdata solutions thanother sectors inwhichdatasetscurationandcountry-by-countryadaptationisneeded.Thisprojectdemonstratesthrough



pilotreplicationsthatsolutionsdevelopedinthissectorareimmediatelyimplementedandadoptedbystakeholdersindifferentcountriesandwithdifferentcontexts.TheEUispursuingthepoliticalaims(1)toreduceCO2emissionsby40%by2030,(2)toreducethenumberofdeathsbycaraccidentsbyhalfby2020and(3)tosupportmethodsformoreeffectiveuseofcarsandautomobilefleets.Thesegoalsstemfromthecircumstancethateveryyear26,000people intheEUdie incaraccidents.600milliontonsofCO2areproduced,andofthatamountamajorproportioniscausedbypoorlymaintainedvehiclesandsubstandardtrafficrouting.Further,logisticsisconsideredasonedriverfortheDigitalSingleMarket (e.g., currently 62% of companies trying to sell online say that too-highparceldeliverycostsareabarrier”).

BioeconomyRelevanceoftheSectorLaunchedandadoptedon13February2012,Europe'sBioeconomyStrategyaddressestheproductionofrenewablebiologicalresourcesandtheirconversionintovitalproductsandbio-energy.ExperiencesfromUSshowthatbioeconomyandspecificallyagriculturecangetasignificantboostfromBigData.InEurope,thissectorhasuntilnowattractedfewlargeICT vendors.Nevertheless, the share of bioeconomy is remarkably large in the nationaleconomyinEUcountries.TheEuropeanbioeconomyisalreadyworthmorethan€2trillionannuallyandemploysover22millionpeople,ofteninruralorcoastalareasandinSmallandMediumSizedEnterprises(SMEs).Farm machines, fishing vessels and forestry machinery in use today collect largequantitiesofdatainnewandpreviouslyunimaginableways.Remoteandlocalsensorsandimagery,andmanyothertechnologies,areallworkingtogethertogivedetailsaboutsoilcontent,marineenvironment,weedsandpests,sunlightandshade,andmanyotherfactors. Analyzing these data can help the farmers, foresters and fishers to adjust theiractivities.Thechallenge is thatapplyingthisdata inpracticerequiresagreatdealofanalysisandsynthesis, bothdataderived from the farm itself and fromothersources.Furthermore,largedatasets-suchasthosecomingfromtheCopernicusearthmonitoringinfrastructure-arebecomingmoreandmoreavailableondifferentlevelsofgranularity,buttheyareheterogonous(insomecasesalsounstructured,hardtoanalyzeanddistributedacrossvarioussectorsanddifferentproviders).NatureofDataAssetsintheSectorThe agriculture sector data assets include Sentinel-1/2 data,machinerymonitoring andsensormeasurements.Asanexample,DataBio,theBigDataPPPLighthouseprojectfocusedon this sector, is planning agriculture pilot implementations expected to utilize 53TBvolumeand197TB/yearvelocityofdata.TheEuropeanCopernicusspaceprogramhascurrentlylauncheditsthirdSentinelsatellitewhosedatawillbeaddedtotheabove.The forestry sector data assets include Sentinel-1/2 forest image data, national forestresource databases (like the Finish Metsään.fi database), forest canopy height models,drainage basin data and aerial image data. Planned forestry pilot implementations inDataBioareexpectedtoutilize12TBvolumeand12TB/yearvelocityofdata.Inaddition,aerial/UAV data offer hyperspectral remote sensing imaging with sample velocities of100GB/h.



Thefisheriessectordataassetsincludeshipoperationalandmotiondata,datafromFADandbuoys,shipenginesensors,VMS,AIS,firstsalenotesfromfisheries’auctions,bigvesselactivitiesasrecordedbyauthories,earthobservationandearthmodels.PlannedfisheriespilotimplementationsinDataBioareexpectedtoutilize9TBvolumeand7TB/yearvelocityofdata.In thetimescale,BigDataexpertsprovidecommonanalytic technologysupport for themaincommonandtypicalBioeconomyapplications/analyticsthatarenowemerging:

• Past:Managingandanalysingdatafromthepast-includingmanydifferentkindofdata sources, i.e. Descriptive analytics and classical query/reporting (in need ofVarietymanagement - andhandlingandanalysisof all of thedata from thepast,including performance data, transactional data, attitudinal data, descriptive data,behavioural data, location-related data, interactional data, from many differentsources).

• Present: Monitoring and real-time analytics - pilot services (in need of Velocityprocessing - and handling of real-time data from the present) - trigging alarms,actuatorsetc.

• Future:Forecasting,PredictionandRecommendationanalytics-services(inneedofVolumeprocessing-andprocessingoflargeamountsofdatacombiningknowledgefromthepastandpresent,andfrommodels,toprovideinsightforthefuture).

ImpactBigDatatechnologiescanespeciallycontributetoimprovingtheprocessesinproductionofbest possible raw materials from agriculture, forestry and fishery for the bioeconomyindustrytoproducefood,energyandbiomaterials.Farmmachines,fishingvessels,forestrymachineryandremoteandproximalsensorscollectlargequantitiesofdata.Largescaledatacollectionandcollationenhancesknowledgetoincreaseperformanceandproductivityinasustainableway.In hisAgenda for Jobs,Growth,Fairness andDemocratic Change, President Junckeridentified 10 key priorities for the EuropeanCommission. The bioeconomy and ICT arecentraltoatleastthreeofthem:

• New Boost for Jobs, Growth and Investment. The innovative bioeconomy is animportantsourceofnewjobs-especiallyatlocalandregionallevel,andinruralandcoastalareas-andtherearebigopportunitiesforthegrowthofnewmarkets,forexampleinbio-fuels,foodandbio-basedproducts

• Resilient Energy Union with a Forward-Looking Climate Change Policy. Europeneeds to diversify its sources of energy and can support breakthroughs in low-carbontechnologieswithcoordinatedresearch.Replacingfossilrawmaterialswithbiological resources is an indispensable component of a forward-looking climatechangepolicy.

• DeeperandFairerInternalMarketwithaStrengthenedIndustrialBase.Innovativebio-basedandfoodindustrieswillcontributeinraisingtheshareofindustryinGDPfrom16%to20%andtocreatingacircular,resource-efficienteconomy.ThefoodanddrinkindustryisalreadythelargestmanufacturingsectorintheEU.



In addition,marine issues and food security are two aspects of the bioeconomywhereEuropecanandshould lead theglobal agendaaspartofPresident Juncker's strategy tomaketheEUastrongerglobalactor.Agriculturalproductivity grows throughbringingnewresources intoproduction (newland, extension of irrigation,or input intensification per hectare) or through raising theproductivityofexistingresources.TheappropriatemeasureofproductivitygrowthinthiscontextisTotalFactorProductivity(TFP)growth,whichisdefinedastheaggregatequantityofoutputsproducedbytheagriculturalsectordividedbytheaggregatequantityofinputsusedtoproducethoseoutputs.TFPgrowthintheEU-27overthepastdecadewasaccordingtoEUDGAGRIadisappointing0.6%perannum(ibid).TheUSDAstatisticsshowedhighernumbers-anaveragegrowthof2.5%varyingbetween4.4%and-0.1%(ibid).BigDatatechnologiesareestimatedtocontributeover30%productivityincrease,whichin5yearperspectivemeansmorethanadoubleannualgrowthrate(USDAnumbers).In forestry, the ratio of value added generated within the forestry and logging sectorcomparedwiththeforestareaavailableforwoodsupplyisoneindicatoroftheproductivity.Theindicatorshowsthatin2011thehighestsharesofvaluegeneratedperforestareaintheEUwereinPortugal(419€/hectare)andthelowestinGreece(18€/hectare).Whenweightedwiththeforestarea,theannualgrowthfortheEUcountrieswithavailabledatahasduring2005-2012beenabout3.5%.Productivityincertaindimensions(mainlycost)canbeatleastdoubledfromthecurrentannualone.Somebenefitswillmaterialiseoveralongertimewithannualincreasesunder1%.Regarding fishery productivity, it is a general trend that the production of the worldfisherieshasstagnatedthelatestdecades.Thisiscausedbythefactthatmostfishstocksareeitheroverexploitedoratitsmaximumyield,andformostspeciesthisisgovernedbythesetquotas.Toincreaseproductivity,BigDatacancontributeinthefollowingareas:

• Betterstockmanagement:Byimprovingthemonitoringofthefishstocks,quotascanbebetteradaptedtoactualsituations.Thiswilldecreasetheriskofcollapse,anditwill help keep the stocks at the size where it is most productive, leading toestimationsof10%increaseinproductivity.

• Improvedenergyefficiency:Theoilconsumptioninthefisheriesdependspartlyonimmediateoperationalchoices,partlyonvesseldesignandpartlyonplanningofthefisheries. Better immediate decision making and planning could lead to 10%decreaseofoilconsumption.

• Improvedmarketadaptation:Thevalueofthefishisverydependentonbothqualityandtiming.Byhelpingthefishingvesseltocatchthecorrectspeciesatthecorrecttime, aproduction increaseof5% is expected.Byalso improvingquality throughdecreasedvesselmovements,another2%increaseindeliveredvalueisexpected.

In addition, Big Data technology providers have the opportunity to expand to newmarkets with new products/services, where, in some sectors, there is very smalladaptation of BigData services and products, offering thus opportunities for very largegrowths. For example, currently there are no Big Data based commercial services foragriculturalprofessionalsinseveralEuropeancountriesandthenumberofBigDatausersinfisheryisnegligible.Inforestry,newBigDataproductsincludeforesthealthmonitoringinnear real time, species classificationbased on temporalbehaviourof each species, orillegalloggingdetection.



EventhoughtheDataBioprojectisthemostprominentexampleofactivitiesinthissectorinthePPP,otherprojectstacklesub-sectorsorspecificusecasesinbio-domains,asitisthecaseofBigDataGrapes,andtosomeextentBigDataOcean.

SmartManufacturingIndustryRelevanceoftheSectorTheManufacturingsectorcontributesperseformorethan15%totheoverallGDPofEU28,additionally mobilizing huge resources from other sectors. Industry 4.0 was born inGermanyin2011asthefourthindustrialrevolution,drivenbytheadoptioninproductionoftheso-calledCyberPhysicalSystems,butverysoonbecametheflagforanyICT-drivenmodernizationofmanufacturingindustry.TheECCommunication“DigitisingEUIndustry”ofApril19th2016,identifiedthreemaingroupsoftechnologiesimplementingIndustry4.0.namelyInternetofThings,BigDataandArtificialIntelligence-DeepLearning,allofthemconcurringinsynergytothedevelopmentofnewsmartproducts(digitalinside),newdigitizedprocessesandnewICT-enabledbusinessmodels.InalmostallEUCountriesandRegions (e.g. the Smart Specialisation Strategy onAdvancedManufacturing and theVanguard Initiative on Efficient and Sustainable Manufacturing), initiatives have beendevelopedtoallowSMEstoaccesstechnologyandknowledgeandtostimulatetheadoptionofIndustry4.0principles.NatureofDataAssetsScenariosinthemanufacturingindustrycanbegroupedinthreemaincategories.SmartFactoryscenariosarerelatedtotheRealWorldofproductionsystemsandplantsalongwiththeirinteractionwiththeDigitalWorld(DigitalTwin).Dataaremostlygeneratedfromthefield by production systems (e.g. robots, machine tools, production lines, conveyors,workplaces,sensors)andgenerallyneedtobeprocessedathighspeed,inordertotaketheproperfactorymanagementdecisions.Typicalapplicationsconcerntheoptimizationoftheproduction,themanagementofwasteandenergy,thezerodefectqualityofproducts,thediagnosisandpredictivemaintenanceaswellasthesafetyandwellbeingattheworkplaceforBlueCollarWorkers.SmartLifecyclescenariosarerelatedtotheproductlifecycleandits phases of ideation, design, manufacturing, operations, maintenance and end-of-life.StructuredDataareoftenenrichedbysentimentsandopinionssometimesalsogatheredfromsocialnetworks.Usually,real timeprocessingperformance isnotrequired,but theamountsofdataareoftenvery largeand inspecificcasesrequirespecializedcomputingarchitectures and resources (HPC). Typical applications concern the modelling andsimulationofproductproperties,thedesignofnewproduct-services,theimplementationofpost-salesservicessuchastrainingandmaintenanceandthesupporttonewbusinessmodels dictated by Sharing and Circular Economy paradigms. Smart Supply Chainscenarios are related to the digitalization of the value chain and involve differentstakeholderssuchassuppliers,distributors,retailers,alllinkedtothemanufacturers.Datatobeinteroperatedarenotjusthuge,butalsoveryheterogeneousandcouldconcerncross-border legal and policy issues. Semantic technologies are often in use to achieveinteroperability.Ownership, confidentialityandprivacy concerns canalsobe relevant inthesescenarios.Typicalapplicationsconcerntheoptimizationofthevarioustiersofsupplychain,thecreationofnon-hierarchicalbusinessecosystems,themanagementofdistributionandlogisticsprocesses,thesynchronizationofonlineandphysicalretail,thedevelopment



ofproperSales&Operationsreplenishmentpoliciesandtheimprovementofconsumers’experienceatthepointofsale.ImpactInitsrecent“TheEuropeanDataMonitoringTool”,IDCreportsaconsiderableexpansionoftheEUDataMarket(60bnEUR)alsoinrelationswiththetotalICTmarket(620bnEUR).Three futurescenarioshavebeendeveloped(HighGrowth,Baseline,Challenge)bringingthecurrent60bnEURto107-80-70bnEURin2020respectively.ManufacturingisalwaysconsideredintoppositionregardingthecurrentsituationandfutureprojectionsoftheDataMarket.Inthecurrent60bnEURscenario,Manufacturingisleadingthegroupofapplicationsby12.8bnEUR,withFinancialandProfessionalservices justbehind. In thefutureBaselinescenario2020(80bnEUR)Manufacturingisregisteringanimportantjumpto17.3bnEURbyconsolidatingitsleadingpositionwithrespecttoFinanceandProfessionalServices.There is thereforenodoubt that Industry4.0and industrialmodernizationareopeninginterestingnewbusinessopportunitiesintheDataMarket,alsointhemostprudentprojections.ActivitiesonSmartManufacturingIndustryinthePPPareverydynamic,withagroupthatholdsperiodicmeetings,workshops,keepsadirectrelationshipwithEFFRA(withwhomBDVA has signed a collaboration agreement) and has produced a white paper). A“Connected Smart Factory” input has also been delivered to the Digitising EuropeanIndustry WG2 on Digital Platforms, as example of dissemination and awareness tocommunities beyond the BDVA. Activities in this community count on the directparticipationofthelighthouseprojectBOOST4.0.

AglimpseontheSMIgrouphistoryTheroadmapoftheSmartManufacturingIndustrygroupin2016followed4majorsteps:(1)engagean internal criticalmass ofmembers participating to the topic (15-20 attendees at AG breakoutsessionsandmorethan60emailsinthelist)(2)gettheapprovalfromtheBoardofDirectorsasofficialsubgroupofTF7applications,(3)agreeamongthestrategyandrationaleforapositionpaperand(4)runaneffectiveinteractiveworkshopattheSummitinordertocollectinputsandgetconsensusfromthecommunityontheR&Itopicswiththefinalaimtoelaborateapositionpaper.

Theactivitiesofthegroupstartedinsummer2016withaseriesofbreakoutsessionsattheActivityGroupMeetingsinSeptemberandOctober.ThesuccessfuloutcomesofthesesessionsbothintermsofattendanceandofexpressionsofinterestresultedtotheBDVABoDapprovalforaTF7subgroup.

ThenewlyestablishedsubgrouporganizedasessionattheValenciaBDVASummitonDecember1st2016withtwosubsequentslotsof90minuteseach.Thefirstpart,attendedbymorethan70people,was dedicated to set the context among all the participants, providing presentations on concreteexperiencesreportedbyBDVAmembers,industrialvisionsandinvestmentsandrelevantreferencescomingfromotherassociation(e.g.EFFRA).Thesecondsession,attendedbynearly50peopleandleveragingontheinteractionandcollaborationamongattendees,hasidentifiedthemosturgentR&IchallengesforfullyadoptingBigDatatechnologiesinManufacturingIndustry.Attendantsweresplitinthreegroups,addressinginturntheidentifiedManufacturingIndustryGrandChallenges:(1)SmartFactory (factory automation and workplace interaction scenarios), (2) Smart Product Lifecycle(Beginning, Middle and End of Life scenarios) and (3) Smart Business Ecosystems (Supply andDistribution chain scenarios). Each one of these three scenarioswas analyzed in relation to thetechnical priorities defined in theBDVA SRIA. As a conclusion of the groupwork, three selectedrapporteurssummarizedandexposedthemainfindingsofthewholeattendance.Thecompletegroupworkdoneonposterswithpost-itswascapturedinpictures,reportedinadocumentandstoredintheBDVAinternaldocumentrepositoryascommonresourceandpredominantlyasthedrivingforceforthepositionpaper.



HealthcareRelevanceoftheSectorThehealthcaresectorcurrentlyaccountsfor8%ofthetotalEuropeanworkforceandfor10%of the EU’sGDP.However, public expenditure on healthcare and long-term care isexpected to increase by one third by 2060. This is primarily due to a rapidly agingpopulation, rising prevalence of chronic diseases and costly developments in medicaltechnology.BigData technologieshave alreadymade impact in fields related tohealthcare:medicaldiagnosisfromimagingdatainmedicine,quantifyinglifestyledatainthefitnessindustry,etc.Nevertheless,thehealthcarehasbeenlaggingintakinguptheBigDatatechnologiesduetoseveralchallenges.NatureofDataAssetsThere are mainly four key verticals generating vast amounts of data within thehealthcare sector: Healthcare providers (e.g. hospitals, GP, laboratories), Insurance,PharmaceuticalsandtheHealthTechsectors.Overthepastdecade,eachverticalhasstartedanalysing its own data sets in order to derive insightswhich are subsequently used toimprovethequalityoftheirproductofferings.Suchsiloedapproachestoderivinginsightsonly fromdata generated by the vertical itself, places a limiton the ability to introduceinnovationsthatcanmakeagenuinedifference.Thishasaseriousimpactontheabilitytomeettheeverincreasingdemandsofthehealthcaresector.Therefore,collecting,integratingandanalysinghealthdataisacomplexandchallengingtaskduetolackofinteroperabilityandharmonizationofdataformats,processingtechniques,datastoragesandtransfersanddifferentlegalframeworks.Asaresult,derivinginsightsandvaluefromtheaggregationofthesedatasetsisnotpossibleatthemoment.Inhealthcare,different typesof informationareavailable fromdifferentsourcessuchaselectronic health care records, patient summaries, genomic and pharmaceutical data,clinical test results, imaging (e.g. x-ray,MRI, etc.), insurance claims, vital signs frome.g.,telemedicine,mobileapps,homemonitoring,on-goingclinicaltrials,real-timesensordata,and informationonwellbeing,behaviourandsocioeconomic indicators.Thisdatacanbebothstructuredandunstructured.Analysinghealthdatacomingfromthedifferentsourcesischallenging.Ontheotherhand,healthdatapresentsvaluableopportunities.Betterclinicaloutcomes,moretailoredtherapeuticresponses,anddiseasemanagementwith improvedqualityoflifeareallappealingaspectsofdatausageinhealth.However, because of the personal and sensitive nature of health data, special attentionneeds tobepaid to legal andethical aspects concerningprivacy.Tounlock itspotential,health(andgenomic)datasharing,withallthechallengesitpresents,isoftennecessarytoensuresuchendeavoursareundertakenresponsibly.Achievingsuchavisionwhichinvolvestheintegrationofsuchdisparatehealthcaredatasets(intermsofdatagranularity,quality,type(e.g.rangingfromfreetext,images,(streaming)sensordata to structureddatasets)posesmajor legal,businessand technical challengesfromadataperspective,intermsofthevolume,variety,veracityandvelocityofthedatasets.Alltheseaspectsleadtotheneedfornewalgorithms,techniquesandapproachestohandlethesechallenges.Besides,neworconsolidatedstandardsareanecessityforsolvingtheproblemsrelatedtodataexchangeandinteroperability.



Asaresult,BigDatamustnotonlybeaddressedfromatechnologyperspective,butalsofocuson the legal andethical issues that come intoplaywhensensitivehealthcaredataneedstobesharedbeyondthetraditionalboundariesofthehealthcareprovideroracrossinternational borders. Furthermore the whole end-to-end value chain in the healthcontinuum needs to be covered starting from Prevention, to Diagnosis, Treatment andeventually Home care where the data generated not onlywithin the boundaries of thehospitalbutalsoinprimarycaresettingsandbypatientsthemselveswillbeutilized.ImpactThekeychallengesthatarerelevanttobetackledinordertoachieveasignificantimpactintheHealthcaresectorare:

• BringingtogetherthekeyplayerswithintheHealthcaresector:HealthcareprovidersandHealthTech,PharmaceuticalandInsuranceindustries.

• Integrating data value chains across Healthcare providers, HealthTech,Pharmaceutical and Insurance industries to derive new, actionable insights toachievebreakthroughquality,accessandyetreducecosts.

• Integratingheterogeneousdatasourceswithverydifferentcharacteristicsintermsof Volume, Variety, Velocity, Veracity and Value, e.g. data from Primary andsecondarycare,EMR,Laboratorydata,Prescription,Imaging,Patientfeedback,Real-timelocationtracking,Patientmonitors(bothhospitalandhome-based),etc.

• BridgingthelanguagegapinherentinEMRsystemsacrossEuropeinordertohaveaunifiedapproachtoperformBigDataanalyticsonhealthcaredatasetsspreadacrossmultipleEuropeancountries.

• Copingwithdifferent legal andethical frameworksacrossEurope so that theBigDataconceptsdevelopedinalighthouseprojectcanberolledoutacrossEurope.

The fusion of healthcare data from multiple sources could take advantage of existingsynergiesbetweendatatoimproveclinicaldecisionsandtorevealentirelynewapproachestotreatingdiseasesandkeepcitizenshealthy.InsightsderivedfromsuchdatageneratedbythelinkingamongEMRdata,vitaldata,laboratorydata,medicationinformation,symptoms(tomention someof these)and their aggregation, evenmorewithdoctornotes,patientdischarge letters, patient diaries, medical publications, namely linking structured withunstructureddata,canbecrucialtodesigncoachingprogrammesthatwouldhelpimprovepeoples’ lifestyles and eventually reduce incidences of chronic disease, medication andhospitalization.Evidence suggests that by improving the productivity of the health care system, publicspendingsavingswouldbelarge,approaching2%ofGDPonaverage in theOECDwhichwouldbeequivalentto€330billioninEuropebasedonGDPfiguresfor2014.TheBigDatatechnologies have the potential to unlock vast productivity bottlenecks and radicallyimprove the quality and accessibility of the healthcare system by disrupting the IronTriangle of Healthcare and allowing simultaneous optimization of all itscomponents-quality,accessandcost,whichisimpossibleatthemoment.TheactivitiesregardingthissectoraredriveninthePPPbyBDVATF.7Healthcaresubgroupand have resulted in a whitepaper entitled “Big Data Technologies in Healthcare”. Theprocess has matchmaking events and workshops in different events. BDVe has had asupportiveroletothecurrentstructuressofarwithoutanespeciallyactiveapproachinthisdomain.However,thefactthatalighthouseprojectisrunningsincethebeginningof2018



(BigMedilytics)willhelptoestablishadditionalsynergieswithotherprojectsthatholdusecasesinthissectorandcommunitiesbeyondthePPP.

TelecommunicationsRelevanceoftheSectorBigDataiscommonlyrecognizedasafundamentalgamechangingsetoftechnologiesacrossmany industrial sectors and societal solutions. The biggest challenge for Big Data is toovercomethefragmentationofthemarketandtoachievecross-fertilizationbetweenmanyestablishedsilos.Telcooperatorscouldserveasplatformbringingdifferentothersectorstogethertoexchangedataandbuildnewbusinessmodels.Thissupportsthemonetizationof data, the digital transformation of vertical industries and the fostering ofentrepreneurship of data. Upcoming technologies like Edge Computing support thistransformation,byprovidingrealtimeinformationtoverticalsattheedge(verticaledgeintelligence), supporting use cases like autonomous driving, industrial internet,transport/logisticsandhealthcare.In the telecommunication sector technologies are evolvingmuch faster than previouslybasedonsoftwarization(e.g.SDN,NFV,5G,IoT,IndustrialInternet,EdgeComputing,Cloud).Thishasaswellimpactonthebusiness/operationalsideandplaysamajorroletosupportthetransformationoftelcooperatorstowardsnewbusinessopportunities.

NatureofDataAssetsintheSectorTelecom operators hold large amounts of data in a variety of locations - central anddistributed databases, as well as OSS, BSS, and CRM systems complemented by datacollectedthruverticals(e.g.IoT,Industry4.0).Thedatahasgreatvalueintermsofcross-fertilizationandanalytics,whichbyfarexceedsthevaluecurrentlyassignedtothedatabyitself.Giventhefactthatdataisakeyasset,thetelecommunicationsectorcouldleverageonitsuniquepositionwithrespecttotheeverincreasingamountofdatasources.Takingthelegalframeworkintoaccount(accountability,transparency,privacy,…)telecomoperatorscanprovidetherighttechnologicalplatformsandtechniquestoensureitsenforcementandsupporttheconceptofPersonalDataSpaces.ImpactTechnical challengeswith respect todata analytics in the telecommunication sector areunsupervisedlearning,deeplearningfornetworks(tosolvedifficulttasksinthenetwork),reinforcementlearning,predictiveanalytics,cognitiveknowledge(generationofvalue,i.e.knowledge, from data), real-time processing of large amounts of data (generating newknowledge, not previously available), and confidential data processing in e.g. cloudenvironments. This underlines the impact of Big Data and Data Analytics from atechnologicalangle,whileitimpactsaswellbusinessmodelstodayandinthefuture.PPPactivitiesinthissectorarebasicallyledbythesubgroupontelecommunications(underTF7),whichhas organized severalworkshops in the context of PPP event and is in theprocessofelaboratingapositionpaper.Someinteractionshavebeenproducedwiththe5GPPP;however,withlittleornoimpactandtheactivityhasnotbeenprioritizedsofar.



MediaRelevanceoftheSectorThe Big Data trend has impacted all industries, including the media industry, as newtechnologiesarebeingdevelopedtoautomateandsimplifytheprocessofdataanalysis,So,Big data is no longer a buzzword in themedia context, but itmust be considered as amainstreamtechnologyespecially for thecontentandnews industry. Inaddition,mediasectorprovidesenormousdataresourcesandtheEuropeancontributionisofleadingroleinthisamountofdata.ThenotionofmediaindustryinthiscontextcoversafullrangeofdifferentindustriesactiveintheMediaMarket:fromTVchannels,tonewspapers,fromLargeBroadcasterstoSMEs.ThesynergybetweenDataSciencetechnologiesandmediaapplicationsisoneofthemostpromisingdimensionofICTapplications,thisisparticularlytrueforSMEs.Thissynergycanbring together competencies from industries and researchdomains,with a strong crossdisciplinaryor interculturaldimensionandapervasive impacton the full value chainofmedia companies, namely: production, distribution, personalisation services andadvertising.Inaddition,abetterandmoreintegratedtechnologyusageinsidenews,mediaandcontentproducerisbecomingurgenttoprotectourdemocracyfromfastgrowingofpopulistsandfakenewsdistribution.DespitethelargegapinthemediaindustriesintermofBigDatacompetencies,accordingtoFinancialTimes,halfofleadersinthesectoralsoseealliances-particularlywithtechnologypartners-asastrategicpriority.Someofthespecificcompetenciesthatwillbeimpactedare(1)videoandaudioanalytics,(2)recommendationsystems,(3)targetingadvertisingand(4)audienceanalyticsandscreening.Big Data and Media means also the birth of new professional specialty, such as DataJournalist. Data journalism is a journalism specialty reflecting the increased role ofnumericaldataintheproductionofinformationinthedigitalera.Itreflectstheincreasedinteractionbetweencontentproducers(journalist)andseveralotherfieldssuchasdesign,computerscienceandstatistics.

NatureofDataAssetsintheSectorThemaindatasourcesofthemediaindustrycovermediacontentassets(videos,photos,text andgraphics),userbehaviourdata,networkeddata, customerdataandmanyotherdatasources.Mediacontentanddatarepresentaboutthe70%oftheInternet’sstoredandshareddata,whichisgrowingexponentially.However,theinterlinkingofthesedatasourcesiscurrentlyonlyemergingandthefullpotentialforexploitationofallthesedatasourcesisnot yet in place. The diversity of European content sources has a huge potential.However,thefragmentationofEuropeanmediaorganisationspreventsthefullexploitationof the European data sources versus the monopolistic approach of Large USA Webcompanies, such asGoogle orFacebook. Themedia related data sources are stored andmanagedinmanydifferentdatasilosacrossEurope.ImpactTheintegrationofalargesetofdiverseandnewinformationcanprovide,indeed,relevantinsightstoaidorganisationsintacklingworld’shardestsocialproblems.However,untilnowthe societal impact of big data has been affected by incomplete, low quality, and even



incorrectdata5.Byallowingthefusingthedatafromdisparatesourcesofinformation,BigData and Media can empower journalists and citizens to comprehend an issue in verycompletewaysandprotectthemfromthefakenewsand“post-true”effects.The use of Big Data inmedia can impact the social good given the peculiar rolemediaindustryplaysatasocietallevelatthecrossroadsofmultiplesetsofdata.Ofcourse,datahastobecollectedinwaysthatmatchourvaluesystemsandrespectethics,privacy,andinformed consent. Big Data applied to Media has an enormous potential to improvesociety-muchlikeotherfundamentaldiscoveriesandtechnologyadvancement,itcangiveus amuchmore accurate and timelier understanding of howour societyworks, so thatdecisionsarebasedonactualfactsratherthandoubtfulhunches.Itisimportantthatpublicmediausesandshowsresultsfromworkingwith“BigData”-toraiseexpectationsofthegeneralpublicacrossthewholeofEuropeofhowdatacanbeused.Whileotherareassuchaseducation,healthandtrafficaresuretochangeifdatawereused,mediaistheareawherethingsareseenbymany,notjustfew.Theimpactofthe“Panamapapers” published in early April shows to some extent how the application of Big Dataanalytics and data capabilities from raw data to understandable narration is utterlyimportant.ActivitiesintheMediasectorinthePPParedrivenbythecorrespondingsubgroupofBDVA.Activityhasbeen limited so fareven thoughsomecommunitybuildingworkshopshavebeen held and discussions have involved Media players. Connections with the NEMcommunityareinplaceandweforecastincreasingactivityinthecomingperiodasaresultof thedevelopmentofrelatedprojects in theprogram, suchasFandango (unfortunatelysomeoftheactivitiesofthesecondwaveofPPPprojectscannotbereportedbecausetheyhaveallstartedinthefirsthalfof2018).

5WorldBankhasidentifiedtheabilitytoproduce,capture,communicateandanalyzebigdataasthefundamentalactiontodrivenew

formsofgrowthandsocialdevelopment,unlockingitsdatathroughitsOpenDataInitiative.

d6.1 dissemination and liaison plan - big data benchmarking€¦ · deliverable d6.1 dissemination...

Documents