Exploration and Visualization in the Web of Big Linked Data: A Survey of the State of the Art

0.16 MB·

by  Nikos Bikakis

Exploration and Visualization in the Web of Big Linked
Data: A Survey of the State of the Art ∗
Nikos Bikakis Timos Sellis
NTUAthens& SwinburneUniv.
ATHENAR.C. ofTechnology
Greece Australia
6
1 ABSTRACT [35]. However,mostoftheseapproachesfailtotakeintoaccount
0
issuesrelatedtoperformanceandscalability.
2 Dataexplorationandvisualizationsystemsareofgreatimportance
Inthiswork,wedescribethemajorrequirementandchallenges
intheBigDataera. Exploringandvisualizingverylargedatasets
n thatshouldbeaddressedbythemodernexplorationandvisualiza-
has become a major research challenge, of which scalability is a
a tionsystems. Additionally,werefertostate-of-the-artapproaches
vitalrequirement. Inthissurvey, wedescribethemajorprerequi-
J
fromtheDatabaseandInformationVisualizationcommunities,which
sites and challenges that should be addressed by the modern ex-
9 attempttohandlesomeofthesechallenges. Further, wedescribe
plorationandvisualizationsystems. Consideringthesechallenges,
2 thesystemsthathavebeendevelopedinthecontextofWoD,and
wepresenthowstate-of-the-artapproachesfromtheDatabaseand
discusstowhichextenttheysatisfythecontemporaryrequirements.
InformationVisualizationcommunitiesattempttohandlethem.Fi-
]
C nally,wesurveythesystemsdevelopedbySemanticWebcommu-
2. CHALLENGES
nityinthecontextoftheWebofLinkedData,anddiscusstowhich
H
extentthesesatisfythecontemporaryrequirements. Mosttraditionalexplorationandvisualizationsystemsoperatein
s. anofflineway,limitedtoaccessingstaticsetsofpreprocesseddata.
c Keywords Additionally,theyrestrictthemselvestodealingwithsmalldataset
[ sizes,whichcanbeeasilyhandledandexploredwithconventional
Visualanalytics,bigdatachallenges,dataexploration,largedatabases,
datamanagementand(visual)explorationstechniques.
1 visualexploration,semanticweb,visualizationtools,scalability
Ontheotherhand,nowadays, theBigDataerahasrealizedthe
v
availabilityofthegreatnumber andvarietyofverylargedatasets
9 1. INTRODUCTION
5 that aredynamic innature. Forexample, most datasourcesoffer
0 The purpose of data exploration and visualization is to offer queryorAPIendpointsforonlineaccessandupdates;inothercases
8 waysforinformationperceptionandmanipulation,aswellasknowl- (e.g.,scientificdatabases),newdataisconstantlyarrived(e.g.,on
0 edgeextractionandinference[68,56].Datavisualization1provides a daily/hour basis). Beyond these, modern systems should oper-
. users with an intuitive means to explore the content of the data, ateonexploratorycontext. Inanexplorationscenario,itcommon
1 identifyinterestingpatterns,infercorrelationsandcausalities,and thatusersareinterestinginfindingsomethinginterestinganduse-
0
supports sense-making activities. Dataexploration andvisualiza- ful without previously know what exactly aresearching for, until
6
tionsystemsareofgreatimportanceintheBigDataera,inwhich the time they identify it. In this case, users perform a sequence
1
thevolumeandheterogeneityofavailableinformationmakeitdif- ofoperations(e.g.,queries),inwhichtheresultofeachoperation
:
v ficultforhumanstomanuallyexploreandanalysedata. determine the formulation of the next operation. Finally, an in-
i Most traditional systems cannot handle the large size of many creasinglylargenumberofdiverseusers(i.e.,differentpreferences,
X
contemporary datasets. Exploring and visualizing large datasets skills,etc.) exploreandanalysedatainaplethoraofdifferentsce-
r hasbecomeamajorresearchchallenge[24,119,55,103,140,49]. narios.
a
Therefore, modern systems have to take into account scalability, Therefore, some of the major challenges that should be dealt
as a vital requirement. Dealing with scalability, modern systems withbymodernsystems,areposedbythe: (1)Largesizeandthe
havetoaddressnumerousissuesrelatedtostorage,access,render- dynamicnatureofdatainconjunctionwiththeexploration-driven
ing/presentation,interaction,etc. setting;and(2)Varietyoftasksandusers.
IntheWebofData(WoD)context,followingtheabundanceof
Linked Data, several recent efforts have offered tools and tech- Large&DynamicDatainExploration-drivenSetting. Oneof
niquesforexplorationandvisualizationinmanydifferentdomains themajorchallengesinexplorationandvisualizationisrelatedto
thesizethatcharacterizes mostcontemporary datasets. Asecond
∗
Thispaperappearsin6thInternationalWorkshoponLinkedWeb challengeisrelatedtotheavailabilityofqueryandAPIendpoints
DataManagement(LWDM2016). foronlinedataaccessandretrieval,aswellasthecaseswherethat
1Throughoutthispaperweusetheterm"visualization"referringto dataisreceivedinastreamfashion.Thelaterposethechallengeof
visualdataexploration.
handlinglargesetsofdatainadynamicsetting,andasaresult,a
preprocessingphase(e.g.,traditionalindexing)isprevented.
(cid:13)c2016, Copyright is withtheauthors. Published intheWorkshopPro- In this respect, modern visualization and exploration systems
ceedingsoftheEDBT/ICDT2016JointConference(March15,2016,Bor- must be able to efficiently and effectively handle billion objects
deaux,France)onCEUR-WS.org(ISSN1613-0073). Distributionofthis dynamicdatasets throughout anexploratory scenario. Therefore,
paperispermittedunderthetermsoftheCreativeCommonslicenseCC-
scalableandefficientdatastructuresandalgorithmshavetobede-
by-nc-nd4.0
LWDM‘16March15,2016,Bordeaux,France veloped. Crucial issues related to storage, access, management,
presentation,interaction(e.g.,pan,zoom,search,drill-down),etc. flyexplorationandanalysismustbecoupledwiththediversityof
over large dynamic datasets have to be handled. Scalability has preferencesandrequirementsposedbydifferentusersandtasks.
becomeamajorchallengeforthemodernsystems. Beyondthese, Therefore,themodernsystemsshouldprovidetheuserwiththe
systemshavetoefficientlyoperateonmachineswithlimitedcom- abilitytocustomizetheexplorationexperiencebasedonherprefer-
putationalandmemoryresources(e.g.,laptops). encesandtherequirementsposedbytheexaminedtask.Forexam-
Ina"conventional" setting(e.g., exploreasmallfragment of a ple,systemsshouldallowtheuserto:(1)organizedataintodiffer-
preprocessed dataset), most of the aforementioned issues can be entways,accordingtothetypeofinformationorthelevelofdetail
handled by the traditional systems that provide database explo- shewishestoexplore(e.g.,[25]);(2)modifyapproximationcrite-
rationandanalysis, suchasTableau2 (previouslyknow asPolaris ria,thresholds,samplingrates,etc. (e.g.,[78]);(3)defineherown
[124]),DEVise[98],Spotfire [3],VisDB[81],Lumira3,QlikView4, operations for data manipulation and analysis (e.g., aggregation,
Datawatch5,etc.However,ina"modern"setting,whenalargepart statistical, filtering functions), etc. Furthermore, systems should
(or the whole) of a billion objects dynamic dataset has to be ex- automatically adjust their parameters, by taking into account the
plored, the aforementioned traditional database-oriented systems environmentsetting(e.g.,screenresolution,memorysize)[74,25,
cannotbeadopted. 73].
In conjunction with performance issues, modern systems have Beyondthepersonalization,modernsystemsshouldprovidemech-
to address challenges related to visual presentation and interac- anismsthatassisttheuserandreducetheeffortneededontheirpart.
tion issues. Particularly, systems should be able to present, as Inthisdirection,severalapproacheshavebeenrecentlydeveloped.
well as, offer ways to "easily" explore large datasets. Handling Inwhatfollows,wementionsomeofthemostcommonones.Sev-
alargenumber ofdataobjectsisachallengingtask; modernsys- eralsystemsassistusersbyrecommendingvisualizationthatseems
temshaveto"squeezeabillionrecordsintoamillionpixels"[119]. tobemoreusefulorcapturesurprisingand/orinterestingdata;e.g.,
Even,inmuchsmallerdatasets,offeringadatasetoverviewisex- [139, 134, 82]. Other approaches help users to discover interest
tremelydifficult; inbothcasesinformationoverloadingisacom- areasinthedataset; bycapturinguser interests,theyguideherto
monissue.Asaslostatedinthevisualinformationseekingmantra: interestingdataparts;e.g.,[37].Finally,inothercasessystemspro-
"overview first, zoom and filter, then details on demand" [118], videexplanationsregardingdatatrendsandanomalies;e.g.,[141].
gainingoverviewiscrucialinthevisualexplorationscenario.Based
on the aforementioned, it follows that a basic requirement of the
3. EXPLORATION &VISUALIZATION
modernsystemsistodevelopmethodsthatprovidesummariesand
abstractionsovertheenormousnumberofdataobjects. SYSTEMS
Inorder totacklebothperformance andpresentationsissues, a
Thissectionreviewsworksrelatedtoexplorationandvisualiza-
large number of systems adopt approximation techniques (a.k.a.
tionintheWoD.Alargenumberofworksstudyingissuesrelated
data reduction techniques) in whichpartial resultsarecomputed.
toWoDvisualexplorationandanalysishavebeenproposedinthe
Considering the existing approaches, most of them are based on:
literature[35,101,4].Inwhatfollows,weclassifytheseworksinto
(1)samplingandfiltering[46,105,2,69,17];or/and(2)aggrega-
the following categories: (1) Browsers and exploratory systems
tion(e.g.,binning, clustering)[42, 25,74,73,97,138,96,1,15,
(Section3.1), (2)Genericvisualizationsystems(Section3.2), (3)
71].Inthisrespect,somemoderndatabase-orientedsystemsadopt
Domain,vocabulary&device-specificvisualizationsystems(Sec-
approximationtechniquesusingquery-basedapproaches(e.g.,query
tion3.3),(4)Graph-basedvisualizationsystems(Section3.4),(5)
translation,queryrewriting)[17,74,73].
Ontologyvisualizationsystems(Section3.5),and(6)Visualization
Inordertoimproveefficiencyseveralsystemsadoptincremental
libraries(Section3.6).
(a.k.a.progressive)techniques.Inthesetechniquestheresults/visual
elementsarecomputed/constructedincrementallybasedonuserin- 3.1 Browsers & Exploratory Systems
teractionorastimeprogresses(e.g.,[123,25]). Numerousrecent
WoD browsers have been the first systems developed for WoD
systemsintegrateincrementalandapproximatetechniques,inthese
utilization and analysis [35, 4]. Similarlyto the traditional ones,
approaches,approximateanswersarecomputedincrementallyover
WoD browsers provide the functionality for link navigation and
progressivelylargersamplesofthedata[46,2,69].
representationofWoDresourcesandtheirproperties;thusenabling
Thedynamicsettingpreventsmodernsystemsfrompreprocessed
browsing and exploration of WoD in amost intuitive way. WoD
thedata.Additionally,itiscommoninexplorationscenariosonlya
browsersmainlyusetabularviewsandlinkstoprovidenavigation
smallfragmentofdatatobeaccessedbytheuser.Inthiscontext,an
overtheWoDresources.
adaptiveindexingapproach[67]isusedin[144],wheretheindexes
Haystack [111] is one of the first WoD browsers, it exploits
are created incrementally and adaptively throughout exploration.
stylesheetsinordertocustomizethedatapresentation. Similarly,
Similarly, in [25] the hierarchy tree is incrementally constructed
Disco6rendersallinformationrelatedtoaparticularRDFresource
based onuser’sinteraction. Finally, insomeapproaches, parallel
as HTML table with property-value pairs. Noadster [113] per-
architecturesareadopted;e.g.,[41,78,77,69].
forms property-based data clustering in order to structure the re-
Tosumup,modernsystemsshouldprovidescalabletechniques
sults. PiggyBank[66]isaWebbrowserplug-in,thatallowsusers
thaton-the-flyeffectively(i.e.,inawaythatcanbeeasilyexplored)
toconvertHTMLcontentintoRDF.LESS[13]allowsuserstocre-
handlealargenumberofdataobjectsoveranexplorationscenario,
atetheir ownWeb-basedtemplatesinorder toaggregateand dis-
usingalimitednumberofresources
playWoD.Tabulator[21]anotherWoDbrowser,additionallypro-
vides maps and timeline visualizations. LENA[87] provides dif-
Variety of Tasks&Users. Therequirement of scalable, on-the-
ferent views of data, following user’s criteria that are expressed
asSPARQLqueries. Visor[110]providesamulti-pivotapproach
2tableau.com forexploringgraphs,allowinguserstoexploremultiplenodesata
3sap-lumira.com time,aswellastoconnectpointsofinterest.Finally,inthecontext
4clickview.com
5datawatch.com 6www4.wiwiss.fu-berlin.de/bizer/ng4j/disco
Table1:GenericVisualizationSystems
System Year DataTypes⋆ Vis.Types⋆⋆ Recomm. Preferences Statistics Sampling Aggregation Incr. Disk Domain App.Type
Rhizomer[30] 2006 N,T,S,H,G C,M,T,TL ✓ generic Web
VizBoard[135,136,109] 2009 N,H C,S,T ✓ ✓ ✓ generic Web
LODWheel[126] 2011 N,S,G C,G,M,P generic Web
SemLens[59] 2011 N S ✓ generic Web
LDVM[29] 2013 S,H,G B,M,T,TR ✓ generic Web
Payola[84] 2013 N,T,S,H,G C,CI,G,M,T,TL,TR generic Web
LDVizWiz[11] 2014 S,H,G M,P,TR ✓ generic Web
SynopsViz[26,25] 2014 N,T,H C,P,T,TL ✓ ✓ ✓ ✓ ✓ ✓ generic Web
VisWizard[131] 2014 N,T,S B,C,M,P,PC,SG ✓ ✓ generic Web
LinkDaViz[129] 2015 N,T,S B,C,S,M,P ✓ ✓ generic Web
ViCoMap[112] 2015 N,T,S M ✓ generic Web
⋆N:Numeric,T:Temporal,S:Spatial,H:Hierarchical(tree),G:Graph(network)
⋆⋆B:bubblechart,C:chart,CI:circles,G:graph,M:map,P:pie,PC:parallelcoordinates,S:scatter,SG:streamgraph,T:treemap,TL:timeline,TR:tree
of faceted browsing, /facet [62], Humboldt [86] and gFacet [57] Rhizomer[30]providesWoDexplorationbasedonaoverview,
providefacetednavigationoverWoDresources. zoomandfilterworkflow. Rhizomeroffersvarioustypesofvisu-
Explorator [7] is a WoD exploratory tool that allows users to alizationssuchasmaps,timelines,treemapsandcharts. VizBoard
browse a dataset by combining search and facets. VisiNav [53] [135,136,109]isaninformationvisualizationworkbenchforWoD
isasystemthatallowsuserstoposeexpressiveexploratory-based buildontopofamashupplatform. VizBoardpresentsdatasetsin
queries. The system is built on top of following concepts: key- a dashboard-like, composite, and interactive visualization. Addi-
wordsearch,objectfocus,pathtraversal,andfacetselection.Infor- tionally,thesystemprovidesvisualizationrecommendations. Pay-
mationWorkbench (IWB)[52] isageneric platformforsemantic ola[84]isagenericframeworkforWoDvisualizationandanaly-
datamanagementofferingseveralback-end(e.g.,triplestore)and sis.Theframeworkoffersavarietyofdomain-specific(e.g.,public
front-endtools.Regardingthefront-end,IWBoffersaflexibleuser procurement) analysis plugins (i.e., analyzers), aswell as several
interfacefordataexplorationandvisualization. Marbles7 formats visualizationtechniques(e.g.,graphs,tables). Inaddition,Payola
RDFtriplesusingtheFresnelvocabulary(avocabularyforrender- offers collaborative features for users to create and share analyz-
ingRDFresourcesasHTML).Also,itretrievesinformationabout ers. In Payola the visualizations can be customized according to
aresourcebyaccessingSemanticWebindexesandsearchengines. ontologiesusedintheresultingdata.
Finally, URI Burner8 is a service which retrieves data about re- TheLinkedDataVisualizationModel(LDVM)[29]providesan
sources.Fortherequestedresources,itgeneratesanRDFgraphby abstract visualization process for WoD datasets. LDVM enables
exploitingexistingontologiesandotherknowledgefromtheWeb. theconnectionofdifferentdatasetswithvariouskindsofvisualiza-
tionsinadynamicway. Thevisualization processfollowsafour
3.2 Generic VisualizationSystems stageworkflow: Sourcedata,Analyticalabstraction,Visualization
InthecontextofWoDvisualexploration,thereisalargenumber abstraction,andView.LDVMconsidersseveralvisualizationtech-
ofgenericvisualizationframeworks,thatofferawiderangeofvi- niques,e.g.,circle,sunburst,treemap,etc. Finally,theLDVMhas
sualizationtypesandoperations. Next,weoutlinethebestknown beenadoptedinseveralusecases[85]. VisWizard[131]isaWeb-
systemsinthiscategory. basedvisualizationsystem, whichexploitsdatasemanticstosim-
InTable1weprovideanoverviewandcompareseveralgeneric plifytheprocessofsettingupvisualizations. VisWizardisableto
visualizationsystems.TheYearcolumnpresentsthereleaseddate. analysemultipledatasetsusingbrushingandlinkingmethods.Sim-
The Data Types column specifies the supported data types. The ilarly,LinkedDataVisualizationWizard(LDVizWiz)[11]provides
Vis.Typescolumnpresentsthetypesofvisualizationsthatarepro- asemi-automaticwayfortheproductionofpossiblevisualization
vided. TheRecomm. column indicates systems that offer recom- for WoD datasets. In a same context, LinkDaViz [129] finds the
mendation mechanisms for visualization settings (e.g., appropri- suitablevisualizationsforagivepartofadataset. Theframework
atevisualizationtype, visualizationparameters). ThePreferences usesheuristicdataanalysisandavisualizationmodelinordertofa-
column captures theabilityof theusers toapply data(e.g., filter, cilitateautomaticbindingbetweendataandvisualizationoptions.
aggregate) or visual (e.g., increase abstraction) operations. The BalloonSynopsis[117]providesaWoDvisualizerbasedonHTML
Statisticscolumn capturestheprovisionof statisticsabout thevi- and JavaScript. It adopts a node-centric visualization approach
sualizeddata. TheSamplingcolumnindicatessystemsthatexploit in a tile design. Additionally, it supports automatic information
techniques based on sampling and/or filtering. The Aggregation enhancement of the local RDF data by accessing either remote
columnindicatessystemsthatexploittechniquesbasedonaggrega- SPARQLendpointsorperformingfederatedqueriesoverendpoints
tion(e.g.,binning,clustering).TheIncr.columnindicatessystems using the Balloon Fusion service [116]. Balloon Synopsis offers
thatadoptincrementaltechniques;i.e.,theresults/visualizationare customizable filters, namely ontology templates, for the users to
computed/generatedbasedonuserinteractionorastimeprogresses. handle and transform (e.g., filter, merge) input data. LODWheel
Finally,theDiskcolumnindicatessystemsthatuseexternalmem- [126]isaWeb-basedvisualizing tool whichcombines JavaScript
ory(e.g.,file,database)toperformoperationsduringruntime(i.e., libraries(e.g.,MooWheel,JQPlot)inordertovisualizeRDFdata
notjustinitiallyloaddatafromdisk). inchartsandgraphs. SemLens[59]isavisualtoolthatcombines
scatterplotsandsemanticlenses,offeringvisualdiscoveryofcor-
7mes.github.io/marbles relations and patterns in data. Objects are arranged in a scatter
8linkeddata.uriburner.com plotandareanalysedusinguser-definedsemanticlenses.ViCoMap
[112] combines WoD statistical analysis and visualization, in a providekeywordsearchfunctionality. TheFiltercolumnindicates
Web-based tool, which offers correlation analysis and data visu- systemsthatprovidemechanismsfordatafiltering. Notethat,Ta-
alizationonmaps. ble2alsoincludestheontologyvisualizationsystems(Section3.5)
Finally, SynopsViz [26, 25] is a Web-based visualization tool thatfollowanode-linkapproach(indicatedbyusingtheterm"on-
built on top of a generic tree-based model. The adopted model tology"intheDomaincolumn).
performs a hierarchical aggregation, allowing efficient personal- RelFinder[58]isaWeb-basedtoolthatoffersinteractivediscov-
izedmultilevelexplorationoverlargedatasets. Inordertoprovide ery and visualization of relationships (i.e., connections) between
scalabilityunderdifferentexplorationscenarios,themodeloffersa selected WoD resources. Fenfire [54] and Lodlive [31] are ex-
methodthatincrementallyconstructsthehierarchybasedonuser’s ploratory tools that allow users to browse WoD using interactive
interaction,aswellasamethodthatenablesdynamicandefficient graphs. StartingfromagivenURI,theusercanexploreWoDby
adaptationofthehierarchytotheuser’spreferences. following the links. LODeX [19] is a tool that generates a rep-
resentative summary of aWoD source. The tool takes as input a
3.3 Domain, Vocabulary & Device-specific SPARQLendpointandgeneratesavisual(graph-based) summary
VisualizationSystems of the WoD source, accompanied by statistical and structural in-
formation of the source. IsaViz [108] allows users to zoom and
Inthissection,wepresentsystemsthattargetvisualizationneeds
navigateovertheRDFgraph, andalsoitoffersseveral"edit"op-
for specific types of data and domains, RDF vocabularies or de-
erations (e.g., delete/add/rename nodes and edges). In the same
vices.
context,graphVizdb[23,22]isbuiltontopofspatialanddatabase
Several systems focuson visualizing and exploring geo-spatial
techniquesofferinginteractivevisualizationoververylarge(RDF)
data. Map4rdf [92] is a faceted browsing tool that enables RDF
graphs. ZoomRDF [142] employs a space-optimized visualiza-
datasetstobevisualizedonanOSMorGoogleMap. Facete[122]
tionalgorithminordertoincreasethenumberofresourceswhich
is an exploration and visualization tool for SPARQL accessible
are displayed. Trisolda [38] proposes a hierarchical RDF graph
data, offering faceted filtering functionalities. SexTant [20] and
visualization. It adopts clustering techniques in order to merge
Spacetime[133]focusonvisualizingandexploringtime-evolving
graph nodes. Paged Graph Visualization (PGV) [36] utilizes a
geo-spatial data. TheLinkedGeoData Browser [121] isafaceted
Ferris-Wheel approach to display nodes with high degree. RDF
browserandeditorwhichisdevelopedinthecontextofLinkedGeo-
graph visualizer [115] adopts a node-centric approach to visual-
Dataproject. Finally,inthesamecontextDBpediaAtlas[132]of-
izeRDFgraphs. Ratherthantryingtovisualizethewholegraph,
fersexplorationovertheDBpediadatasetbyexploitingthedataset’s
nodes of interest (i.e., staringnodes) are discovered bysearching
spatialdata. Furthermore,inthecontextoflinkeduniversitydata,
overnodeslabels;thentheusercaninteractivelynavigateoverthe
VISUalizationPlayground(VISU)[6]isaninteractivetoolforspec-
graph. RDF-Gravity9 visualizes RDF and OWL data. It offers
ifyingandcreatingvisualizationsusingthecontentsoflinkeduni-
filtering, keyword search and editing the graph layout. Also, the
versitydatacloud. Particularly,VISUoffersanovelSPARQLin-
nodescanbedisplayedindifferentcolorsandshapesbasedontheir
terfaceforcreatingdatavisualizations.Queryresultsfromselected
RDFtypes.Adifferentapproachhasbeenadoptedin[127],where
SPARQLendpointsarevisualizedwithGoogleCharts.
samplingtechniqueshavebeenexploited. Finally,Gephi[15]isa
AvarietyofsystemstargetmultidimensionalWoDmodelledwith
generic toolthat offersseveral visualizationand analysisfeatures
theDataCubevocabulary. CubeViz[43,114]isafacetedbrowser
overgraphdata.
forexploringstatisticaldata. Thetoolprovidesdatavisualizations
usingdifferenttypesofcharts(i.e.,line,bar,column,areaandpie). 3.5 OntologyVisualizationSystems
ThePayolaDataCubeVocabulary[60]adoptstheLDVMstages
The problems of ontology visualization and exploration have
[29] in order to visualize RDF data described by the Data Cube
been extensively studied in several research areas (e.g., biology,
vocabulary. ThesametypesofchartsasinCubeVizareprovided
chemistry). Inwhatfollowswefocusongraph-basedontologyvi-
in this tool. The OpenCube Toolkit [75] offers several tools re-
sualizationsystemsthathavebeendeveloped intheWoDcontext
latedtostatisticalWoD.Forexample,OpenCubeBrowserexplores
[47, 40, 51, 91, 80]. In most systems, ontologies are visualized
RDFdatacubesbypresentingatwo-dimensionaltable. Addition-
following the node-link paradigm [100, 99, 64, 104, 27, 45, 65,
ally,theOpenCubeMapViewoffersinteractivemap-basedvisual-
94,5,89,125]10,11. Ontheotherhand,CropCircles[137]usesa
izationsofRDFdatacubesbased ontheirgeo-spatial dimension.
geometriccontainment approach, representingtheclasshierarchy
TheLinkedDataCubesExplorer(LDCE)[79]allowsuserstoex-
asasetofconcentriccircles. Furthermore,hybridsapproachesare
ploreandanalysestatisticaldatasets. Finally,[106]offersseveral
adoptedinotherworks.Knoocks[88]combinescontainment-based
mapandchartvisualizationsofdemographic,socialandstatistical
andnode-linkapproaches. Inthiswork,ontologiesarevisualized
linkedcubedata.
asnestedblockswhereeachblockisdepictedasarectanglecon-
Regarding device-specific systems, DBpedia Mobile [18] is a
tainingasub-branchshownastreemap.Finally,OntoTrix[14]and
location-awaremobileapplicationforexploringandvisualizingDB-
NodeTrix[61]usenode-linkandadjacencymatrixrepresentations.
pedia resources. Who’s Who [32] is an application for exploring
andvisualizinginformationfocusingonseveralissuesthatappear 3.6 VisualizationLibraries
in the mobile environment. For example, the application consid-
Finally,thereisavarietyofJavascriptlibrarieswhichallowWoD
erstheusabilityanddataprocessingchallengesrelatedtothesmall
visualizations to be embedded in Web pages. Sgvizler [120] isa
displaysizeandlimitedresourcesofthemobiledevices.
JavaScript wrapper for visualizing SPARQL results. Sgvizler al-
3.4 Graph-based VisualizationSystems lowsuserstospecifySPARQLSelectqueriesdirectlyintoHTML
elements.SgvizlerusesGoogleChartstogeneratetheoutput,offer-
A large number of systems visualize WoD datasets adopting a
ingnumerousvisualizationstypessuchascharts,treemaps,graphs,
graph-based(a.k.a.,node-link)approach[102].InTable2wepro-
vide an overview and compare several graph-based visualization 9semweb.salzburgresearch.at/apps/rdf-gravity
systems. Table2isstructuredinasimilarwaytoTable1. Addi- 10protegewiki.stanford.edu/wiki/OntoGraf
tionally, in this table the Keyword column indicates systems that 11protegewiki.stanford.edu/wiki/OWLViz
Table2:Graph-basedVisualizationSystems
System Year Keyword Filter Sampling Aggregation Incr. Disk Domain App.Type
RDF-Gravity9 2003 ✓ ✓ generic Desktop
IsaViz[108] 2003 ✓ ✓ generic Desktop
RDFgraphvisualizer[115] 2004 ✓ generic Desktop
GrOWL[89] 2007 ✓ ✓ ✓ ontology Desktop
NodeTrix[61] 2007 ✓ ontology Desktop
PGV[36] 2007 ✓ ✓ generic Desktop
Fenfire[54] 2008 generic Desktop
Gephi[15] 2009 ✓ ✓ ✓ generic Desktop
Trisolda[38] 2010 ✓ ✓ ✓ generic Desktop
Cytospace[127] 2010 ✓ ✓ ✓ ✓ ✓ generic Desktop
FlexViz[45] 2010 ✓ ✓ ontology Web
RelFinder[58] 2010 generic Web
ZoomRDF[142] 2010 ✓ ✓ ✓ generic Desktop
KC-Viz[104] 2011 ✓ ontology Desktop
LODWheel[126] 2011 ✓ ✓ generic Web
GLOW[64] 2012 ✓ ✓ ontology Desktop
Lodlive[31] 2012 ✓ generic Web
OntoTrix[14] 2013 ✓ ✓ ontology Desktop
LODeX[19] 2014 ✓ ✓ generic Web
VOWL2[100,99] 2014 ontology Web
graphVizdb[23,22] 2015 ✓ ✓ ✓ ✓ generic Web
timelines,etc.Visualbox[50]providesanenvironmentwhereusers tionsbasedonuserinteraction. Asaresult,eachtime,onlyapart
can build and debug SPARQL queries in order to retrieve WoD; oftheexamineddatasetneedstobeloadedinmainmemory.
then, a set of visualization templates is provided to visualize re- Thegraph-basedexplorationandvisualizationsystemsarepre-
sults. Visualbox uses several visualization libraries like Google sentedinTable2. ThesesystemsareofgreatimportanceinWoD,
ChartsandD3[28],offering14visualizationtypes. duetothegraphstructureoftheRDFdatamodel.Althoughseveral
systemsoffersamplingoraggregationmechanisms,mostofthese
systems load the whole graph in main memory. Given the large
4. DISCUSSION
memoryrequirementsofgraphlayoutalgorithmsinordertodrawa
Inthissectionwediscusstowhichextentthesystemsdeveloped largegraph,thecurrentWoDsystemsarerestrictedtohandlesmall
intheWoDcontextfulfilledthenowadaysrequirements,focussing sizedgraphs.
onperformanceandscalabilityissues,availabilityofpersonalized Inordertobeabletohandlelargegraphs,modernWoDsystems
servicesfacilitiesforassistingusersthroughexploration. should adopt more sophisticated techniques similar to those pro-
As previously mentioned, most of WoD exploration and visu- posed by the information visualization community. Particularly,
alizationsystemsdonothandleissuesrelatedtoperformanceand state-of-the-art systems for exploring large graphs utilize hierar-
scalability. Theybasicallyadopttraditionaltechniquesinorderto chical aggregation approaches where the graph is recursively de-
handlesmallsetsofdata. composedintosmallersub-graphs(inmostcasesusingclustering
AswecanobservefromTable1,genericsystemssupportseveral and partitioning) that form a hierarchy of abstraction layers [93,
typesofdata(e.g.,numeric,temporal,graph,spatial)andprovide 10, 95, 9, 8, 1, 143, 12, 15, 71, 130]. Other approaches adopt
aplethoraofvisualizationtypes. Additionally,anincreasingnum- edgebundlingtechniqueswhichaggregategraphedgestobundles
ber of recent systems (e.g., LinkDaViz, Vis Wizard, LDVizWiz, [48, 44, 107, 90, 34, 63]. Beyond hierarchical approaches, WoD
LDVM)focusonprovidingrecommendationmechanisms. Partic- systemsshouldalsoconsiderdisk-basedimplementations,suchas
ularity,thesesystemsmainlyrecommendthemostsuitablevisual- [22,1,72,127,130].
izationtechniquebyconsideringthetypeofinputdata. Tosumup,WoDcommunityshouldconsiderscalabilityandper-
Regardingvisual scalability, aswecan seeinTable1, none of formance as vital requirements for the development of the future
thesystems,withtheexceptionsofSynopsVizandVizBoardcases, exploration and visualization systems. Handing large datasets is
adopt approximation techniques (i.e., sampling/filtering, aggrega- crucial in the Big Data era. Therefore, in what follows we sum-
tion).Hence,theexistingapproachesassumethatalltheexamined marize some possible directions for the future WoD exploration
dataobjectscanbepresentedonthescreenandhandledbytradi- andvisualizationsystems.Approximationtechniquessuchassam-
tionalvisualizationtechniques.Duetothisassumption,thecurrent plingandaggregationthathavebeenwidelyusedinsystemsfrom
systemsrestricttheirapplicabilitytosmallsetsofdata. database and information visualization communities, have to be
Inconjunctionwiththelimitedvisualscalability,mostoftheex- adopted and adjusted to WoD data and requirements. Systems
istingsystems(exceptforSynopsViz)donotexploitexternalmem- should be integrated with disk structures, retrieving data dynam-
oryduringruntime.Particularly,theyinitiallyloadalltheexamined ically during runtime. Also caching and prefetching techniques
objectsinmainmemory,assumingthatthemainmemoryislarge maybeexploited; e.g., [128, 76, 70, 16, 33, 83, 39]. Datastruc-
enough. AnalternativeapproachisadoptedbytheSynopsVizsys- turesandindexesshouldbedevelopedfocusingonWoDtasksand
tem, which incrementally retrieves data and generates visualiza-
data,suchasNanocubes[96]inthecontextofspatio-temporaldata [22] N.Bikakis,J.Liagouris,M.Krommyda,G.Papastefanatos,
exploration, and HETree [25] in numeric and temporal datasets. andT.Sellis.TowardsScalableVisualExplorationofVery
Finally, considering users’ perspective, beyond visualization rec- LargeRDFGraphs.InESWC,2015.
ommendations,modernWoDsystemsshouldprovidemoresophis- [23] N.Bikakis,J.Liagouris,M.Krommyda,G.Papastefanatos,
andT.Sellis.graphVizdb: AScalablePlatformfor
ticatedmechanismsthatcaptureusers’preferencesandassistthem
InteractiveLargeGraphVisualization.InICDE,2016.
throughoutlargedataexplorationandanalysistasks.
[24] N.BikakisandG.Papastefanatos.VisualExplorationand
AnalyticsofBigData:ChallengesandApproaches,2016.
5. REFERENCES
[25] N.Bikakis,G.Papastefanatos,M.Skourla,andT.Sellis.A
[1] J.Abello,F.vanHam,andN.Krishnan.ASK-GraphView: HierarchicalAggregationFrameworkforEfficient
ALargeScaleGraphVisualizationSystem.TVCG,12(5), MultilevelVisualExplorationandAnalysis,2015.Techn.
2006. Rep.,http://arxiv.org/abs/1511.04750.
[2] S.Agarwal,B.Mozafari,A.Panda,H.Milner,S.Madden, [26] N.Bikakis,M.Skourla,andG.Papastefanatos.
andI.Stoica.BlinkDB:QuerieswithBoundedErrorsand rdf:SynopsViz-AFrameworkforHierarchicalLinkedData
BoundedResponseTimesonVeryLargeData.InEuroSys, VisualExplorationandAnalysis.InESWC,2014.
2013. [27] T.Boinski,A.Jaworska,R.Kleczkowski,andP.Kunowski.
[3] C.Ahlberg.Spotfire:AnInformationExploration Ontologyvisualization.InITNGSWUI,2010.
Environment.SIGMODRecord,25(4),1996. [28] M.Bostock,V.Ogievetsky,andJ.Heer.D3Data-Driven
[4] F.Alahmari,J.A.Thom,L.Magee,andW.Wong. Documents.TVCG,17(12),2011.
EvaluatingSemanticBrowsersforConsumingLinkedData. [29] J.M.Brunetti,S.Auer,R.García,J.Klímek,and
InADC,2012. M.Necaský.FormalLinkedDataVisualizationModel.In
[5] H.Alani.TGVizTab:AnOntologyVisualisationExtension iiWAS,2013.
forProtege.InWorkshoponVisualizingInformationin [30] J.M.Brunetti,R.Gil,andR.García.FacetsandPivoting
KnowledgeEngineering,2003. forFlexibleandUsableLinkedDataExploration.In
[6] M.Alonen,T.Kauppinen,O.Suominen,andE.Hyvönen. InteractingwithLinkedDataWorkshop,2012.
ExploringtheLinkedUniversityDatawithVisualization [31] D.V.Camarda,S.Mazzini,andA.Antonuccio.LodLive,
Tools.InESWC,2013. exploringthewebofdata.InI-SEMANTICS,2012.
[7] S.F.C.Araújo,D.Schwabe,andS.D.J.Barbosa. [32] A.E.Cano,A.Dadzie,andM.Hartmann.Who’sWho-A
ExperimentingwithExplorator:aDirectManipulation LinkedDataVisualisationToolforMobileEnvironments.
GenericRDFBrowserandQueryingTool.InVisual InESWC,2011.
InterfacestotheSocialandtheSemanticWeb,2009. [33] S.Chan,L.Xiao,J.Gerth,andP.Hanrahan.Maintaining
[8] D.Archambault,T.Munzner,andD.Auber.Grouse: interactivitywhileexploringmassivetimeseries.InIEEE
Feature-Based,SteerableGraphHierarchyExploration.In VAST,2008.
EuroVis,2007. [34] W.Cui,H.Zhou,H.Qu,P.C.Wong,andX.Li.
[9] D.Archambault,T.Munzner,andD.Auber.GrouseFlocks: Geometry-BasedEdgeClusteringforGraphVisualization.
SteerableExplorationofGraphHierarchySpace.TVCG, TVCG,14(6),2008.
14(4),2008. [35] A.DadzieandM.Rowe.ApproachestovisualisingLinked
[10] D.Archambault,T.Munzner,andD.Auber.Tugging Data:Asurvey.SemanticWeb,2(2),2011.
GraphsFaster:EfficientlyModifyingPath-Preserving [36] L.Deligiannidis,K.Kochut,andA.P.Sheth.RDFdata
HierarchiesforBrowsingPaths.TVCG,17(3),2011. explorationandvisualization.InWorkshopon
[11] G.A.AtemezingandR.Troncy.Towardsalinked-data CyberInfrastructure: InformationManagementineScience,
basedvisualizationwizard.InCOLD,2014. 2007.
[12] D.Auber.Tulip-AHugeGraphVisualizationFramework. [37] K.Dimitriadou,O.Papaemmanouil,andY.Diao.
InGraphDrawingSoftware.2004. Explore-by-Example: AnAutomaticQuerySteering
[13] S.Auer,R.Doehring,andS.Dietzold.LESS– FrameworkforInteractiveDataExploration.InSIGMOD,
Template-BasedSyndicationandPresentationofLinked 2014.
Data.InESWC,2010. [38] J.DokulilandJ.Katreniaková.UsingClustersinRDF
[14] B.Bach,E.Pietriga,andI.Liccardi.VisualizingPopulated Visualization.InAdvancesinSemanticProcessing,2009.
OntologieswithOntoTrix.IJSWIS,9(4),2013. [39] P.R.Doshi,E.A.Rundensteiner,andM.O.Ward.
[15] M.Bastian,S.Heymann,andM.Jacomy.Gephi:AnOpen PrefetchingforVisualDataExploration.InDASFAA,2003.
SourceSoftwareforExploringandManipulatingNetworks. [40] M.Dudás,O.Zamazal,andV.Svátek.Roadmappingand
InICWSM,2009. NavigatingintheOntologyVisualizationLandscape.In
[16] L.Battle,R.Chang,andM.Stonebraker.Dynamic EKAW,2014.
PrefetchingofDataTilesforInteractiveVisualization, [41] A.Eldawy,M.Mokbel,andC.Jonathan.HadoopViz:A
2015.TechnicalReport. MapReduceFrameworkforExtensibleVisualizationofBig
[17] L.Battle,M.Stonebraker,andR.Chang.Dynamic SpatialData.InICDE,2016.
reductionofqueryresultsetsforinteractivevisualizaton.In [42] N.ElmqvistandJ.Fekete.HierarchicalAggregationfor
BigData,2013. InformationVisualization:Overview,Techniques,and
[18] C.BeckerandC.Bizer.ExploringtheGeospatialSemantic DesignGuidelines.TVCG,16(3),2010.
WebwithDBpediaMobile.J.WebSem.,7(4),2009. [43] I.Ermilov,M.Martin,J.Lehmann,andS.Auer.Linked
[19] F.Benedetti,L.Po,andS.Bergamaschi.AVisual OpenDataStatistics:CollectionandExploitation.In
SummaryforLinkedOpenDatasources.InISWC,2014. KnowledgeEngineeringandtheSemanticWeb,2013.
[20] K.Bereta,C.Nikolaou,M.Karpathiotakis,K.Kyzirakos, [44] O.Ersoy,C.Hurter,F.V.Paulovich,G.Cantareiro,and
andM.Koubarakis.SexTant:VisualizingTime-Evolving A.Telea.Skeleton-BasedEdgeBundlingforGraph
LinkedGeospatialData.InISWC,2013. Visualization.TVCG,17(12),2011.
[21] T.Berners-Lee,Y.Chen,L.Chilton,D.Connolly, [45] S.Falconer,C.Callendar,andM.-A.Storey.A
R.Dhanaraj,J.Hollenbach,A.Lerer,andD.Sheets. VisualizationServicefortheSemanticWeb.InKnowledge
Tabulator:ExploringandAnalyzinglinkeddataonthe EngineeringandManagementbytheMasses.2010.
SemanticWeb.InSWUI,2006.
[46] D.Fisher,I.O.Popov,S.M.Drucker,andM.C.Schraefel. [72] J.F.R.Jr.,H.Tong,A.J.M.Traina,C.Faloutsos,and
TrustMe,I’mPartiallyRight:IncrementalVisualization J.Leskovec.GMine:ASystemforScalable,Interactive
LetsAnalystsExploreLargeDatasetsFaster.InCHI,2012. GraphVisualizationandMining.InVLDB,2006.
[47] B.Fu,N.F.Noy,andM.-A.Storey.EyeTrackingtheUser [73] U.Jugel,Z.Jerzak,G.Hackenbroich,andV.Markl.Faster
Experience-AnEvaluationofOntologyVisualization VisualAnalyticsthroughPixel-PerfectAggregation.
Techniques.SemanticWebJournal,2015. PVLDB,7(13),2014.
[48] E.R.Gansner,Y.Hu,S.C.North,andC.E.Scheidegger. [74] U.Jugel,Z.Jerzak,G.Hackenbroich,andV.Markl.
MultilevelAgglomerativeEdgeBundlingforVisualizing VDDA:automaticvisualization-drivendataaggregationin
LargeGraphs.InPacificVis,2011. relationaldatabases.VLDBJ,2015.
[49] P.Godfrey,J.Gryz,andP.Lasek.InteractiveVisualization [75] E.Kalampokis,A.Nikolov,P.Haase,R.Cyganiak,
ofLargeDataSets,2015.TechnicalReport. A.Stasiewicz,A.Karamanou,M.Zotou,D.Zeginis,
[50] A.Graves.CreationofVisualizationsBasedonLinked E.Tambouris,andK.A.Tarabanis.ExploitingLinkedData
Data.InWIMS,2013. CubeswithOpenCubeToolkit.InISWC,2014.
[51] F.Haag,S.Lohmann,S.Negru,andT.Ertl.OntoViBe:An [76] A.Kalinin,U.Çetintemel,andS.B.Zdonik.Interactive
OntologyVisualizationBenchmark.InVISUAL,2014. DataExplorationUsingSemanticWindows.InSIGMOD,
[52] P.Haase,M.Schmidt,andA.Schwarte.TheInformation 2014.
WorkbenchasaSelf-ServicePlatformforLinkedData [77] A.Kalinin,U.Çetintemel,andS.B.Zdonik.Searchlight:
Applications.InCOLD,2011. EnablingIntegratedSearchandExplorationoverLarge
[53] A.Harth.VisiNav:Asystemforvisualsearchand MultidimensionalData.PVLDB,8(10),2015.
navigationonwebdata.J.WebSem.,8(4),2010. [78] N.Kamat,P.Jayachandran,K.Tunga,andA.Nandi.
[54] T.Hastrup,R.Cyganiak,andU.Bojars.BrowsingLinked DistributedandInteractiveCubeExploration.InICDE,
DatawithFenfire.InWWW,2008. 2014.
[55] J.HeerandS.Kandel.InteractiveAnalysisofBigData. [79] B.KämpgenandA.Harth.OLAP4LD-AFrameworkfor
ACMCrossroads,19(1),2012. BuildingAnalysisApplicationsOverGovernmental
[56] J.HeerandB.Shneiderman.InteractiveDynamicsfor Statistics.InESWC,2014.
VisualAnalysis.Commun.ACM,55(4),2012. [80] A.Katifori,C.Halatsis,G.Lepouras,C.Vassilakis,and
[57] P.Heim,T.Ertl,andJ.Ziegler.FacetGraphs:Complex E.G.Giannopoulou.Ontologyvisualizationmethods-a
SemanticQueryingMadeEasy.InESWC,2010. survey.ACMComput.Surv.,39(4),2007.
[58] P.Heim,S.Lohmann,andT.Stegemann.Interactive [81] D.KeimandH.-P.Kriegel.VisDB:DatabaseExploration
RelationshipDiscoveryviatheSemanticWeb.InESWC, usingMultidimensionalVisualization.IEEEComputer
2010. GraphicsandApplications,14(5),1994.
[59] P.Heim,S.Lohmann,D.Tsendragchaa,andT.Ertl. [82] A.Key,B.Howe,D.Perry,andC.R.Aragon.VizDeck:
SemLens:visualanalysisofsemanticdatawithscatterplots Self-organizingDashboardsforVisualAnalytics.In
andsemanticlenses.InI-SEMANTICS,2011. SIGMOD,2012.
[60] J.Helmich,J.Klímek,andM.Necaský.VisualizingRDF [83] H.A.Khan,M.A.Sharaf,andA.Albarrak.DivIDE:
DataCubesUsingtheLinkedDataVisualizationModel.In efficientdiversificationforinteractivedataexploration.In
ESWC,2014. SSDBM,2014.
[61] N.Henry,J.Fekete,andM.J.McGuffin.NodeTrix:a [84] J.Klímek,J.Helmich,andM.Necaský.Payola:
HybridVisualizationofSocialNetworks.TVCG,13(6), CollaborativeLinkedDataAnalysisandVisualization
2007. Framework.InESWC,2013.
[62] M.Hildebrand,J.vanOssenbruggen,andL.Hardman. [85] J.Klímek,J.Helmich,andM.Necaský.UseCasesfor
/facet:ABrowserforHeterogeneousSemanticWeb LinkedDataVisualizationModel.InLDOW,2015.
Repositories.InISWC,2006. [86] G.KobilarovandI.Dickinson.Humboldt:Exploring
[63] D.Holten.HierarchicalEdgeBundles:Visualizationof LinkedData.InLDOW,2008.
AdjacencyRelationsinHierarchicalData.TVCG,12(5), [87] J.KochandT.Franz.LENA-BrowsingRDFDataMore
2006. ComplexThanFoaf.InISWC,2008.
[64] W.Hop,S.deRidder,F.Frasincar,andF.Hogenboom. [88] S.KriglsteinandR.Motschnig-Pitrik.Knoocks:New
UsingHierarchicalEdgeBundlestovisualizecomplex VisualizationApproachforOntologies.InConferenceon
ontologiesinGLOW.InACMSAC,2012. InformationVisualisation,2008.
[65] A.Hussain,K.Latif,A.Rextin,A.Hayat,andM.Alam. [89] S.Krivov,R.Williams,andF.Villa.GrOWL:Atoolfor
ScalableVisualizationofSemanticNetsusingPower-Law visualizationandeditingofOWLontologies.J.WebSem.,
Graphs.AMIS,8(1),2014. 5(2),2007.
[66] D.Huynh,S.Mazzocchi,andD.R.Karger.PiggyBank: [90] A.Lambert,R.Bourqui,andD.Auber.WindingRoads:
ExperiencetheSemanticWebInsideYourWebBrowser. RoutingEdgesintoBundles.CGF,29(3),2010.
InISWC,2005. [91] M.Lanzenberger,J.Sampson,andM.Rester.Visualization
[67] S.Idreos,M.L.Kersten,andS.Manegold.Database inOntologyTools.InCISIS,2009.
Cracking.InCIDR,2007. [92] A.d.Leon,F.Wisniewki,B.Villazón-Terrazas,and
[68] S.Idreos,O.Papaemmanouil,andS.Chaudhuri.Overview O.Corcho.Map4rdf-FacetedBrowserforGeospatial
ofDataExplorationTechniques.InSIGMOD,2015. Datasets.InUsingOpenData:policymodeling,citizen
[69] J.Im,F.G.Villegas,andM.J.McGuffin.VisReduce:Fast empowerment,datajournalism,2012.
andResponsiveIncrementalInformationVisualizationof [93] C.Li,G.Baciu,andY.Wang.ModulGraph:
LargeDatasets.InBigData,2013. Modularity-basedVisualizationofMassiveGraphs.In
[70] P.Jayachandran,K.Tunga,N.Kamat,andA.Nandi. VisualizationinHighPerformanceComputing,2015.
CombiningUserInteraction,SpeculativeQueryExecution [94] T.LiebigandO.Noppens.OntoTrack:Asemantic
andSamplingintheDICESystem.PVLDB,7(13),2014. approachforontologyauthoring.J.WebSem.,3(2-3),2005.
[71] J.F.R.Jr.,H.Tong,J.Pan,A.J.M.Traina,C.T.Jr.,and [95] Z.Lin,N.Cao,H.Tong,F.Wang,U.Kang,andD.H.P.
C.Faloutsos.LargeGraphAnalysisintheGMineSystem. Chau.DemonstratingInteractiveMulti-resolutionLarge
TKDE,25(1),2013. GraphExploration.InICDMWorkshops,2013.
[96] L.D.Lins,J.T.Klosowski,andC.E.Scheidegger. [121] C.Stadler,J.Lehmann,K.Höffner,andS.Auer.
NanocubesforReal-TimeExplorationofSpatiotemporal LinkedGeoData: Acoreforawebofspatialopendata.
Datasets.TVCG,19(12),2013. SemanticWeb,3(4),2012.
[97] Z.Liu,B.Jiang,andJ.Heer.imMens:Real-timeVisual [122] C.Stadler,M.Martin,andS.Auer.Exploringthewebof
QueryingofBigData.CGF,32(3):421–430,2013. spatialdatawithfacete.InWWW,2014.
[98] M.Livny,R.Ramakrishnan,K.S.Beyer,G.Chen, [123] C.D.Stolper,A.Perer,andD.Gotz.ProgressiveVisual
D.Donjerkovic,S.Lawande,J.Myllymaki,andR.K. Analytics:User-DrivenVisualExplorationofIn-Progress
Wenger.DEVise:IntegratedQueryingandVisual Analytics.TVCG,20(12),2014.
ExplorationofLargeDatasets.InSIGMOD,1997. [124] C.StolteandP.Hanrahan.Polaris:ASystemforQuery,
[99] S.Lohmann,S.Negru,F.Haag,andT.Ertl.VOWL2: AnalysisandVisualizationofMulti-Dimensional
User-OrientedVisualizationofOntologies.InEKAW,2014. RelationalDatabases.InInfoVis,2000.
[100] S.Lohmann,S.Negru,F.Haag,andT.Ertl.Visualizing [125] M.D.Storey,N.F.Noy,M.A.Musen,C.Best,R.W.
OntologieswithVOWL.SemanticWebJournal,2015. Fergerson,andN.A.Ernst.Jambalaya: aninteractive
[101] N.MarieandF.L.Gandon.SurveyofLinkedDataBased environmentforexploringontologies.InIUI,2002.
ExplorationSystems.InIESD,2014. [126] M.Stuhr,D.Roman,andD.Norheim.LODWheel-
[102] S.Mazumdar,D.Petrelli,K.Elbedweihy,V.Lanfranchi, JavaScript-basedVisualizationofRDFData.InCOLD,
andF.Ciravegna.Affectivegraphs:Thevisualappealof 2011.
LinkedData.SemanticWeb,6(3),2015. [127] S.Sundara,M.Atre,V.Kolovski,S.Das,Z.Wu,E.I.
[103] K.Morton,M.Balazinska,D.Grossman,andJ.D. Chong,andJ.Srinivasan.Visualizinglarge-scaleRDFdata
Mackinlay.SupporttheDataEnthusiast:Challengesfor usingSubsets,Summaries,andSamplinginOracle.In
Next-GenerationData-AnalysisSystems.PVLDB,7(6), ICDE,2010.
2014. [128] F.Tauheed,T.Heinis,F.Schürmann,H.Markram,and
[104] E.Motta,P.Mulholland,S.Peroni,M.d’Aquin,J.M. A.Ailamaki.SCOUT:PrefetchingforLatentFeature
Gómez-Pérez,V.Mendez,andF.Zablith.ANovel FollowingQueries.PVLDB,5(11),2012.
ApproachtoVisualizingandNavigatingOntologies.In [129] K.Thellmann,M.Galkin,F.Orlandi,andS.Auer.
ISWC,2011. LinkDaViz-AutomaticBindingofLinkedDatato
[105] Y.Park,M.J.Cafarella,andB.Mozafari. Visualizations.InISWC,2015.
Visualization-AwareSamplingforVeryLargeDatabases. [130] C.Tominski,J.Abello,andH.Schumann.CGV-An
InICDE,2016. InteractiveGraphVisualizationSystem.Computers&
[106] I.Petrou,M.Meimaris,andG.Papastefanatos.Towardsa Graphics,33(6),2009.
methodologyforpublishingLinkedOpenStatisticalData. [131] G.Tschinkel,E.E.Veas,B.Mutlu,andV.Sabol.Using
eJournalofeDemocracy&OpenGovernment,6(1),2014. SemanticsforInteractiveVisualAnalysisofLinkedOpen
[107] D.Phan,L.Xiao,R.B.Yeh,P.Hanrahan,andT.Winograd. Data.InISWC,2014.
FlowMapLayout.InInfoVis,2005. [132] F.Valsecchi,M.Abrate,C.Bacciu,M.Tesconi,and
[108] E.Pietriga.IsaViz:aVisualEnvironmentforBrowsingand A.Marchetti.DBpediaAtlas:MappingtheUncharted
AuthoringRDFModels.InWWW,2002. LandsofLinkedData.InLDOW,2015.
[109] J.PolowinskiandM.Voigt.VISO:AShared,Formal [133] F.ValsecchiandM.Ronchetti.Spacetime:aTwo
KnowledgeBaseAsaFoundationforSemi-automatic DimensionsSearchandVisualisationEngineBasedon
InfovisSystems.InCHI,2013. LinkedData.InSEMAPRO,2014.
[110] I.O.Popov,M.M.C.Schraefel,W.Hall,andN.Shadbolt. [134] M.Vartak,S.Madden,A.G.Parameswaran,and
ConnectingtheDots:AMulti-pivotApproachtoData N.Polyzotis.SEEDB:AutomaticallyGeneratingQuery
Exploration.InISWC,2011. Visualizations.PVLDB,7(13),2014.
[111] D.A.QuanandR.Karger.Howtomakeasemanticweb [135] M.Voigt,S.Pietschmann,L.Grammel,andK.Meißner.
browser.InWWW,2004. Context-awareRecommendationofVisualization
[112] P.RistoskiandH.Paulheim.VisualAnalysisofStatistical Components.IneKNOW,2012.
DataonMapsusingLinkedOpenData.InESWC,2015. [136] M.Voigt,S.Pietschmann,andK.Meißner.A
[113] L.Rutledge,J.vanOssenbruggen,andL.Hardman. Semantics-Based,End-User-CenteredInformation
MakingRDFpresentable:integratedglobalandlocal VisualizationProcessforSemanticWebData.InSemantic
semanticWebbrowsing.InWWW,2005. ModelsforAdaptiveInteractiveSystems.2013.
[114] P.E.R.Salas,F.M.D.Mota,K.K.Breitman,M.A. [137] T.D.WangandB.Parsia.CropCircles:TopologySensitive
Casanova,M.Martin,andS.Auer.PublishingStatistical VisualizationofOWLClassHierarchies.InISWC,2006.
DataontheWeb.Int.J.SemanticComputing,6(4),2012. [138] H.Wickham.Bin-Summarise-Smooth:AFrameworkfor
[115] C.Sayers.Node-centricRDFGraphVisualization,2004. VisualisingLargeData.Technicalreport,2013.
TechnicalReportHPLaboratories. [139] K.Wongsuphasawat,D.Moritz,A.Anand,J.D.Mackinlay,
[116] K.Schlegel,F.Stegmaier,S.Bayerl,M.Granitzer,and B.Howe,andJ.Heer.Voyager:ExploratoryAnalysisvia
H.Kosch.BalloonFusion:SPARQLRewritingbasedon FacetedBrowsingofVisualizationRecommendations.
UnifiedCo-ReferenceInformation.InDESWeb,2014. TVCG,22(1),2016.
[117] K.Schlegel,T.Weißgerber,F.Stegmaier,C.Seifert, [140] E.Wu,L.Battle,andS.R.Madden.TheCaseforData
M.Granitzer,andH.Kosch.BalloonSynopsis:AModern VisualizationManagementSystems.PVLDB,7(10),2014.
Node-CentricRDFViewerandBrowserfortheWeb.In [141] E.WuandS.Madden.Scorpion:ExplainingAwayOutliers
ESWC,2014. inAggregateQueries.PVLDB,6(8),2013.
[118] B.Shneiderman.TheEyesHaveIt:ATaskbyDataType [142] K.Zhang,H.Wang,D.T.Tran,andY.Yu.ZoomRDF:
TaxonomyforInformationVisualizations.InIEEE semanticfisheyezoomingonRDFdata.InWWW,2010.
SymposiumonVisualLanguages,1996. [143] M.Zinsmaier,U.Brandes,O.Deussen,andH.Strobelt.
[119] B.Shneiderman.ExtremeVisualization:Squeezinga InteractiveLevel-of-DetailRenderingofLargeGraphs.
BillionRecordsintoaMillionPixels.InSIGMOD,2008. TVCG,18(12),2012.
[120] M.G.Skjæveland.Sgvizler:AJavaScriptWrapperforEasy [144] K.Zoumpatianos,S.Idreos,andT.Palpanas.Indexingfor
VisualizationofSPARQLResultSets.InESWC,2012. interactiveexplorationofbigdataseries.InSIGMOD,
2014.