BIGDATAANDPRIVACY:ATECHNOLOGICALPERSPECTIVE
26
Dataintegrationisdifferentiatedfromdatafusioninthatintegrationmorebroadlycombinesdatasetsand
retainsthelargersetofinformation.Indata fusion, thereisusuallyareductionorreplacementtechnique.Data
fusionisfacilitatedbydatainteroperability,theabilityfortwosystemstocommunicateandexchangedata.
Data
fusionanddataintegrationarekeytechniquesforbusinessintelligence.Retailersareintegratingtheir
online,in‐store,andcatalogsalesdataba ses tocreatemorecompletepicturesoftheircustomers.Williams‐
Sonoma,forexample,hasintegratedcustomerdatabases withinformationon60millionhouseholds.Variables
includinghouseholdincome,housingvalues,and
numberofchildrenaretracked.Itisclaimedthattargeted
emailsbasedonthisinformationyieldtento18timestheresponserateofemailsthatarenottargeted.
73
Thisis
asimpleillustrationofhowmoreinformationcanleadtobetterinferences.Techniquesthatcanhelpto
preserveprivacyareemerging.
74
Thereisagreatamountofinteresttodayinmulti‐sensordatafusion.
75
Thebiggesttechnicalchallengesbeing
tackledtoday,generallythroughdevelopmentofnewandbetteralgorithms,relatetodataprecision/resolution,
outliersandspuriousdata,conflictingdata,modality(bothheterogeneousandhomogeneousdata)and
dimensionality,datacorrelation,dataalignment,associationwithindata,centralizedvs.decentralized
processing,operationaltiming,andtheabilityto
handledynamic vs.staticphenomena.Privacyconcernsmay
arisefromsensorfidelityandprecisionaswellascorrelationofdatafrommultiplesensors.Asinglesensor’s
outputmightnotbesensitive,butthecombinationfromtwoormoremayraiseprivacyconcerns.
3.2.3Imageandspeechrecognition
Image‐andspeech‐recognitiontechnologiesareabletoextractinformation,insomelimitedcasesapproaching
humanunderstanding,frommassivecorpusesofstillimages,videos,andrecordedorbroadcastspeech.
Urban‐sceneextractioncanbeaccomplishedusingavarietyofdatasourcesfromphotosandvideostoground
basedLiDAR(aremote
‐sensingtechniqueusinglasers).
76
Inthegovernmentsector,citymodelsarebecoming
vitalforurbanplanningandvisualization.Theyareequallyimportantforabroadrangeofacademicdisciplines
includinghistory,archeology,geography,andcomputer‐graphicsresearch.Digitalcitymodelsarealsocentralto
popularconsumermappin g andvisualizationapplicationssuchasGoogleEarth
andBingMaps,aswellasGPS‐
enablednavigationsystems.
77
Sceneextractionis anexampleoftheinadvertentcaptureofpersonal
informationandcanbeusedfordatafusionthatrevealspersonalinformation.
Facial‐recognitiontechnologiesarebeginningtobepracticalincommercialandlaw‐enforcementapplications.
78
Theyareabletoacquire,normalize,andrecognizemovingfacesindynamicscenes.Real‐timevideosurve illance
withsingle‐camerasystems(andsomewithmulti‐camerasystems,whichcanbothrecognizeobjectsand
analyzeactivity)hasawidevarietyofapplicationsinbothpublicandprivateenvironments,suchashomeland
73
Manyika,J.etal.,“BigData:Thenextfrontierforinnovation,competition,andproductivity,”McKinseyGlobalInstitute,
2011.
74
Navarro‐Arriba,G.andV.Torra,"Informationfusionindataprivacy:Asurvey,"InformationFusion,13:4,2012,pp.235‐
244.
75
Khaleghi,B.etal.,"Multisensordatafusion:Areviewofthestate‐of‐the‐art,"InformationFusion,14:1,2013,pp.28‐44.
76
Lam,J.,etal.,"Urbansceneextractionfrommobilegroundbasedlidardata,"Proceedingsof3DPVT,2010.
77
Agarwal,S.,etal.,"BuildingRomeinaday,"CommunicationsoftheACM,54:10,2011,pp.105‐112.
78
WorkshoponFrontiersinImageandVideoAnalysis,NationalScienceFoundation,FederalBureauofInvestigation,
DefenseAdvancedResearchProjectsAgency,andUniversityofMarylandInstituteforAdvancedComputerStudies,January
28‐29,2014.http://www.umiacs.umd.edu/conferences/fiva/