web visualization - tools and approaches for information visualization on the web
DESCRIPTION
The Web, once solely a repository for texts and images, is evolving into a comprehensive platform where any kind of community - scientific, media, entertainment, etc - can publish knowledge. Knowledge is derived from information, thus, in a scientific area it is of utmost importance to make the information available as fast as possible. Along with the web evolution, web development tasks have also evolved. Nowadays, it is easier and faster to develop, test and deploy novel applications. Therefore, it is essential for any researcher to know the existing set of tools that are most adequate to his work. Information visualization is one of the research areas where this awareness is more relevant.TRANSCRIPT
WebVisualization
Tools&ApproachesforInformationVisualizationontheWeb
TópicosAvançadosemEngenhariaInformática2ProgramaDoutoralemEngenhariaInformática
UniversidadedeAveiro
PEDROLOPES
2
Abstract
TheWeb, once solely a repository for texts and images, is evolving into a comprehensive
platformwhereanykindofcommunity‐scientific,media,entertainment,etc‐canpublish
knowledge.Knowledgeisderivedfrominformation,thus, inascientificareait isofutmost
importance to make the information available as fast as possible. Along with the web
evolution,webdevelopment tasks have also evolved.Nowadays, it is easier and faster to
develop,testanddeploynovelapplications.Therefore, it isessential foranyresearcherto
knowtheexistingsetoftoolsthataremostadequatetohiswork.Informationvisualization
isoneoftheresearchareaswherethisawarenessismorerelevant.Anyresearcher,atsome
pointofhiswork,needstopublishinformationonlinerelyingonsomekindofvisualization
technique.Thisreportshowssomeofthemostmodernapproachesandtoolsavailable.The
purposeistomakeresearchersawareofthesetools,enablingtheirusageinscientificwork
andsimplifyingonlinepublishingtasks.
3
TableofContents
Abstract ...................................................................................................................... 2
TableofContents........................................................................................................ 3
1.Introduction............................................................................................................ 5
2.History .................................................................................................................... 7
3.Web2.0 ................................................................................................................. 10
3.1Technologies.................................................................................................... 10
3.1.1Flash ................................................................................................................ 11
3.1.2Silverlight......................................................................................................... 11
3.1.3AJAX ................................................................................................................ 11
3.2Approaches...................................................................................................... 12
3.2.1Textualcontentvisualization .......................................................................... 12
3.2.2Maps ............................................................................................................... 14
3.2.3InteractiveTimelines....................................................................................... 15
3.2.4Mindmaps ....................................................................................................... 16
3.3Tools................................................................................................................ 17
3.3.1Flash ................................................................................................................ 17
3.3.2Silverlight......................................................................................................... 19
3.3.3AJAX ................................................................................................................ 20
3.3.4Summary ......................................................................................................... 25
4.Requirements ....................................................................................................... 26
4.1UserRequirements .......................................................................................... 26
4.2Interface .......................................................................................................... 27
5.Conclusion ............................................................................................................ 28
References................................................................................................................ 29
4
5
1.Introduction
WorldWideWeb is evolving to become the primarymedia for knowledge exchange. The
Internetisnowthemainplatformforelectronicallydeliveredcontent,whetheritisscientific
information,news,entertainmentorsimplepersonalstories.Webevolutionledtothebirth
ofWeb2.0.Withit,anewwaveofapplicationsappeared,empoweringtheuserwithcontrol
overtheweb.Thisshiftincontrolresultedinseveralnewhotapplicationsthatenableusers
topublishtheirownpersonalcontent[1].
With this next generation applications, developers are no longer stuck with desktop
applications and their inconvenient: hard to test, slow to deploy, restricted user scope,
complex maintenance. Now, they can create web applications that resemble desktop
applications in several levels, especially onusability, user interface andavailable features.
Data and information visualization has been one of the subjects of this shift to web
applications.Thisraisedanewclassofchallenge,mostlyduetoresourceconsumptionand
complex web architectures. Even before Web2.0 or even Google, there were several
challenges inweb‐baseddatavisualization [2].With theevolutionof the Internet, sohave
thechallengeschanged.
The focus is now divided in three main challenges: moving complex desktop
visualization applications to a web environment, improve existing user interfaces by
reducingusabilitycomplexities[3,4]andcreatingpowerful–butsimple–datavisualization
components thatcanbeeasilyusedbyanydeveloper, likeGeoBoost [5] for instance.This
reportapproaches the latter issueand tries tomakeageneraloverviewovervisualization
componentsthatcanbehelpfultoeverydaydevelopersandresearchers.
Considering that the web ecosystem is currently the main information exchange
platform, developers and researchers need to publish information online and, in several
cases, largedatasetshavetobepublished inanattractive format.Several frameworksare
beingdeveloped in the scienceworld to facilitate thisprocess, likeWVFA [6] for instance.
However, this report is focused on more widely used scenarios and state‐of‐the‐art
technologies. Therefore, it contains descriptions about the main used technologies for
developing next generationweb applications and a set of tools that can ease the task of
creatingandinteractivetimelineorpresentinggeographicinformation.
6
The following section contains a small historical note about the evolution on web
information visualization. Next, there is the core of this report focused on technologies,
techniques and tools to implement those techniques efficiently. At last there is a general
requirements sectionwith some topics thatwebdevelopers should take in accountwhen
designinganewwebapplication.
7
2.History
Visualization has only been applied recently in web environments. In the beginning, the
Internet was confined to text browsers without any kind of graphic enhancement.
ViolaWWW [7] (Figure 1), in 1993, was the first web browser with support for graphics,
tables,stylesheetsandplugins.Thepossibilitiesthatwereenabledwiththepowertoinclude
visual information in a web page was immense.With this, developers started to include
tables,imagesandempoweredthedevelopmentofstandardslikeCascadeStylingSheets–
CSS–andECMAScript–Javascript.
Figure1‐ViolaWWWbrowsershowingpagewithimages
Whenmaturityintheinclusionofvisualinformationwasreacheditwastimetoinclude
traditional desktop visualization components in web pages. Circa 1997, Java applets
appeared,enablingthecreationofapplicationsthatweredeployedintheclient;thus,being
8
able to implement a new set of interface and visualization features. Along with these
improvements,developersstartedtheworkontheinclusionof3Dgraphics–VRML–inweb
pages[8].TheinclusionofVRMLcontentinwebpagesisnotatrivialtaskanditrequiresa
client‐side plugin for proper functionality. Hence, the users need to install a VRML
applicationinordertoseethethree‐dimensionaldatadisplayedonline.Javaappletswerea
more successful approach. Javawasmorewidely installed thanVRMLand, additionally, it
allowedthedevelopmentofsoftwarewithahigherlevelofcomplexity,insteadofthesimple
VRMLmodels.Evenwithalltheadvances,VRMLisstillusedinseveralfieldsforcreating3D
environmentsinthebrowser[9,10].Javaappletusageisdecreasingwiththeappearanceof
novel client‐side technologies. Someof thebest Java‐based informationvisualization tools
(mixedwithtext‐mining)areJonathanHarris’WeFeelFine1(Figure2)andUniverse2.These
applications gather knowledge that is scattered through several blogs or RSS feeds and
processesit,creatingastonishingvisualresults.However,withtheadventofWeb2.0,anew
set of technologies gained significance in web development. Adobe Flash3 Microsoft
Silverlight4andAJAXarethenewwaveoftechnologiesthatempowerthewebecosystem.
Figure2‐WeFeelFinesampleinterface
1WeFeelFine:http://www.wefeelfine.org2Universe:http://universe.daylife.com3AdobeFlash:http://www.adobe.com/products/flash4MicrosoftSilverlight:http://www.silverlight.net
9
Summarizing, the key point is thatwehavewitnessed a shift in themain purposeof
webdatavisualization.About10yearsago,thegoalwastoadaptdesktop‐basedtechniques
to web pages. This increased the web applications weight and resulted in a complex
deploymentmodel.Nowadays,themainpurposeistousenovelwebtechnologiestooffer
desktop‐like solutions to users.With higher bandwidths and cross‐browser/cross‐platform
technologies,itispossibletocreatewebapplicationsthatresembledesktopapplicationsin
terms of appearance, resources and available features. However, deploying visualization
applicationsinthewebhasincreasedthecomplexityofexistingvisualizationissues[11]and
addedanotherlevelofissuesrelatedtotheamountofinformationthathastobeobtained
andprocessedinreal‐timebyclient‐side–browser‐based–frameworks.
10
3.Web2.0
After the “dot‐com boom”, the Internet needed an orientation change. The exponential
increaseonthenumberofweb‐basedenterprisesledtoasectorcrisis.Thechangecamein
theformofWeb2.0[12].ThemainWeb2.0purposeistogiveusercontrolovercontent.The
typical Web2.0 business model consists in developing profitable concepts that can be
maintainedbyusers ‐ Figure 3. Themajority of the Internet contentdoesnot come from
specialized technicians. Nowadays, the content comes from end users. Applications like
blogs,micro‐blogs,socialnetworks,medianetworks,socialbookmarksormeta‐applications
like mashups or workflows are currently the main trend in web applications. These
applicationsdependonuserstocreatecontentandpublishitonline.
Figure3‐Web1.0vsWeb2.0
3.1Technologies
Aspreviouslymentioned,wehavethreetoptechnologies for thecreationofRich Internet
Applicationsthatrequireexcellentuserinterfaceanddatavisualizationfeatures.Thissection
enclosesabriefdescriptiononeachofthem.Next,therearepracticalexamplesofmodern
visualizationapproachesimplementedwiththesetechnologies.
11
3.1.1Flash
Theoldestone isAdobeFlash. Flashwas initially created fordesigners and it allowed the
creation of high‐quality and visually stimulating interfaces. Recently, Adobe has opted to
updatethenewerversionstoamuchmoredeveloper‐orientedversion.Thenowdesignated
FlashPlatformempowersdevelopersinthecreationofvisuallyricherandmorestructurally
complexapplications,combiningserver‐sidewithclient‐sidedevelopment.Despiterequiring
aplugin,itisalmostcompletelycross‐platformand,inthelastiterations,allowsthedisplay
ofcomplex3Dmodelsandtheprocessingofhugedatasets.Withthis,itiseasiertointegrate
dataandgenerategraphsortoconceivenovelvisualizationstructures.
3.1.2Silverlight
Flash’smaincompetitorisMicrosoftSilverlight.Silverlighthasalmostthesamefeaturesas
Flash but is more developer oriented. Being created for developers it enables the faster
creation and deployment of new components. Currently, it has a growing database of
visualization components ranging from calendars to maps, from timelines to bar charts.
Microsoft’s support is giving Silverlight an important adoption push, especially when it
comestoapplicationsthatneedtodisplaylargeamountsofdatainaninteractiveinterface
thatiscapableoffollowingseveralusabilitygoodpractices.
Themain Silverlight problem is the same as Flash: both require the installation of a
plugin to work. This means that users without software installation permissions or
administrativeaccessmaynotbeabletoviewthedevelopedcomponentsintheirbrowsers.
3.1.3AJAX
AJAXstandsforAsynchronousJavascriptandXMLanditisthemainalternativetoFlashand
Silverlight. AJAX relies especially in Javascript. Javascript has evolved to become the only
dynamic language, supported by browsers, which can interact with the HTML Document
ObjectModel.ThismakesAJAXcross‐browserandcross‐platform.Italsomeansthatitdoes
notrequirethatanykindofpluginginstallation.
The main idea behind AJAX is to take advantage of asynchronous data exchanges
between the client and the server, thus removing web page refreshes. Combining
12
asynchronous requests, XML and Javascript, developers can program their applications to
change a specific component of the page with real‐time server‐obtained data, without
having to reload the page. This makes the applications faster and more responsive by
removing“whitescreens”betweenuserclicks.
3.2Approaches
Datavisualizationinterfacecanbeuniqueintermsofcontent,elegance,beautyand,above
all, description. Histograms, tables, bar or pie graphs are used daily in the most diverse
projects.However,thereareotherwaysofconveyingamessagetouserseffectivelythana
simple histogram. This small section intends to introduce modern data visualization
techniquesandhowtheycanbecreatedusedthepresentedtechnologies.
Despite the fact that there innumerous approaches to display data, the key factor is
creativity. It is imperative to be creative and test new smart ideas of displaying large
datasets. Inaddition, it isalso important to remember thatmostof the times the simpler
methodsarethemosteffectiveones.Thisadvice isvalidnotonlytodatavisualizationbut
alsotothedesignofhuman‐computerinterfaces.
3.2.1Textualcontentvisualization
TheappearanceofRSS feedsallowedusers to syndicate contentandpublish it in formats
thatcanbeused,autonomously,byanyapplication, leveragedthecreationofseveralnew
visualization tools. These visualization tools mostly provide timeline displays of news or
graphicaldisplayofrelevanttopics.
Newsmaps (Figure4) is aparticular implementationof the traditional treemapview,
developed in Flash. This informationvisualizationapproachorganizes thehierarchical tree
organizationinseveralconnectedsquares,representingnewsandrelationsbetweennews.
Thisapproachcanbeveryusefulwhenonewantstodisplaylargetextualcontentthatis,in
somemanner,related.Similardatavisualizationmechanismscanbeimplementedtogather
information concerning a particular topic or interest area instead of simple reading the
13
internationalnewspapersRSS feeds.Apracticalexamplecouldbethedisplayofamedical
ontology,containingdiseases,phenotypesandsymptoms.
ThesameteambehindWeFeelFinecreatedtheUniversalDaylifeviewer(Figure5)to
accessworldwidenewsdatasearchingfortopicsanddisplayingtheminavirtualuniverse.
Theusermaythenwatchtheskyandsearchforstarsandconstellationswhicharedynamic
relations obtained from the provided input. This viewer is a complex Java applet and is a
remarkableachievementinthequalityofthesekindsofapplications.
Figure4‐Newsmapdatavisualizationinterface5
Figure5‐UniversalDaylifeviewer6
5Newsmap:http://newsmap.jp
14
3.2.2Maps
Advancesinthetechnologymadetheexistenceofworldmapswithdetailedsatelliteimages
possible. There are several map APIs that allow developers to create new layers of
informationontopofthegeographicalmaps.Thismeansthatitiseasierandfastertocreate
visualizationapplicationsdisplayinggeographical‐relateddata.Figure6showshowwecan
use the Google Maps API and connect it with the YouTube API to create a visualization
applicationthatshowsallthecountriesthatparticipatedintheEurovision2009contestand
enablestheuserstoviewthevideosofthesongsinthecontest.
TheseveralmapsAPIsallowconnectionstoanykindofapplicationandthedisplayof
informationrelatedtoanyarea.
6UniversalDaylife:http://universe.daylife.com
Figure6‐GoogleMapsEurovisionContestviewer
15
3.2.3InteractiveTimelines
Timelines[13,14]traditionallydisplaydataorderedinaspecificdatescale.However,static
timelines are not useful to present large amounts of information. Any of thementioned
technologiesallowsthecreationofinteractivetimelinesthatmayincludeameshofcontent
from separate areas and gathered from distinct sources. Figure 7 shows an interactive
timeline,developedbyBBC,displayingthemostrelevanteventsinBritishHistory.
Figure7‐BritishHistoryinteractivetimeline7
TheVoyagenewsreader(Figure8)displaysdatacontentinanimprovedtimeline.The
content is highlightedwith focus changes that create a depth experience in the browser.
Userscannavigateinthetimeline:thenearestlayeristhemostrecentandbygoingdeeper
they viewoldernews.Onceagain, this kindof interactive timeline canbeused in several
distinctscenariosthatrequireinformationpresentationinsomekindofsortablescale.
7BBCBritishHistoryinteractivetimeline:http://bit.ly/pnRur
16
Figure8‐Voyagenewsreadinginterface8
3.2.4Mindmaps
Mindmaps [15] are data visualizations that present several connections and relations in a
waysimilartosubwaymaps.Usinginteractivemindmaps,userscanselectasingleelement
andview/navigateinsideitsrelations,withoutlosingcontextinformationontheremaining
elementsofthemap.Staticmindmapsarepoorerbuttheyprovideamoregenericoverview
on larger datasets. Figure 9 shows amindmap containing theweb search trends in 2007
relatedlikeasubwaymap.
8VoyageRSSNewsReader:http://rssvoyage.com
17
Figure9‐WebTrendsmapin20079
3.3Tools
To use the presented technologies, developers do not need to start from scratch and
developanewsetofcomponentsthatallowthemtocreatemodernwebvisualizationtools.
Thereareseveral toolkitsavailablethatease,significantly,developers’ tasksandallowthe
creationofsolutionstailoredtoeachspecificproject.
3.3.1Flash
The Flash Platform is a success case in merging complex data visualization with fully
functional applications. Demandbase10 (Figure 10), ModelMetrics11 or Nasdaq OMX12
representstate‐of‐the‐artFlash‐basedapplicationswhereinformationvisualizationplaysthe
keyrole.Theseapplicationsdealwithseveraldatabasesandprocessmillionsofrecords, in
real‐time,beforepresentingthemtotheusers.
9WebTrends2007:http://informationarchitects.jp/start10Demandbase:http://www.adobe.com/flashplatform/apps/demandbase11ModelMetrics:http://www.adobe.com/flashplatform/apps/model_metrics12NasdaqOMX:http://www.adobe.com/flashplatform/apps/nasdaq
18
Figure10‐Demandbaseinterface
To create smaller customized solutionone canuse the Flare13ActionScript collection.
ThiscollectioncontainsvariousActionScriptscriptsthatcanbeusedwiththefreeFlexSDK14
tocreateweb‐basedvisualizationapplications.Theavailablescriptsprovideaneasywayto
deploy circular or bar graphs, trees, stacks or timelines (among others) online. Figure 11
shows an example of a complex bar chart created using Flare. This example contains a
multitudeofcategoriesandevaluatedelementsthataredisplayed,automatically,inasingle
compositegraph.
13PrefuseFlare:http://flare.prefuse.org14AdobeFlex:http://www.adobe.com/products/flex
19
Figure11‐PrefuseFlarebarexample
3.3.2Silverlight
Descry15 and several Telerik components16 are perfect examples of Silverlight’s data
visualizationcapabilities.Withthemitispossibletoeasilycreatenovelapplicationsthatcan
merge data gathered from several sources, presenting it to users in a captivating user
interface.Figure12showsapracticalDescryexample.ItcontainsdatafromUSA’s50states
regardingobesepopulationpercentage.Insteadoftraditionalbarcharts,Descry’sapproach
usesT‐shirtsrepresentingeachstatethatarecoloredandenlargedaccordingtothestate’s
obesity parameters, creating a richer visualization interface that users can interact and
analyzeinreal‐time.
15Descry:http://www.visitmix.com/Lab/Descry16TelerikChart:http://www.telerik.com/products/silverlight/chart.aspx
20
Figure12‐DescryvisualizationoftheobesepopulationintheUnitedStatesofAmerica17
3.3.3AJAX
To develop applications using AJAX, one can use any web application technology in the
server‐side.Applicationscanhave thebackgroundwith simpleHTML,XML,ASP.NET, Java,
PHP, Pythonor any other server‐side controller language. Regarding the client‐side, there
aremany frameworks available for composing the typical richuser interface that remarks
Web2.0applications.ThemorewidelyusedAJAXframeworksare:jQuery18,script.aculo.us19,
Prototype20andDojo21.
These frameworkswerebuilt forgenericpurposes.Despite the fact that theyarenot
focusedonsolvingvisualizationissues,therearediversevisualizationcomponentsthatcan
beeasily implemented.Tocreatestate‐of‐the‐artwebdatavisualizationapplicationsthere
areothergroupsofmorefocusedframeworks.Thefollowingparagraphsdescribethebest
Javascriptvisualizationtoolkitsandweb‐basedvisualizationapplications.
The first example comes from the worldly known MIT. MIT’s SIMILE Project22 has
severalteamsworkingondatavisualizationtools.Thisisanongoingresearchproject;hence,
17DescryObesity:http://www.visitmix.com/labs/descry/theobesityepidemic18jQuery:http://jquery.com19Script.aculo.us:http://script.aculo.us20PrototypeJavaScript:http://www.prototypejs.org21DojoToolkit:http://www.dojotoolkit.org22SIMILEProject:http://simile.mit.edu
21
thereareonlya fewstablewidgetsavailable23. Timeline (Figure13)andTimeplotwidgets
are ideal for interactive timeline creation and the display of advanced time‐related
information.
Figure13‐SIMILINETimelineexample
The Javascript InformationVisualizationToolkit24 – JIT– is simple Javascript library to
createadvancedinformationvisualizationapplications.Thistoolkitallowstheeasycreation
and online display of mainly interactive trees and graphs. The focus is given to the easy
creationofvisualizationsdisplayingrelationalinformation.Figure14showsahyperbolictree
containingarelationmapoftheelementsofthePearlJamband.Userscanselectanyband
elementandview,inrealtime,therelationstheyhavewithinthemusicindustry.
23SIMILEWidgets:http://www.simile‐widgets.org24JIT:http://www.thejit.org
22
Figure14‐JITHyperbolicrelationtree
The presented frameworks are good for some specific purposes. However, for more
genericpurposes,thereistheGoogleVisualizationAPI25.Thisextensiveframework,created
andsupportedbyGoogle,containsseveralusefulcomponentsrangingfrommapstotables,
from word clouds to interactive charts. Most of the available components are also
integratedinJavaAJAXGoogleWebToolkit26framework.Beingaframeworkdevelopedby
Google, there is extremely good support and the quality of this framework cannot be
questioned.UsingGoogleVisualizationAPIitispossibletoimplementsomeofthemodern
visualization techniquesmentioned previously. Themore interesting feature in this API is
that the source datasets can be easily obtained via Javascript. This means that one can
25GoogleVisualizationAPI:http://code.google.com/apis/visualization26GoogleWebToolkit:http://code.google.com/webtoolkit
23
dynamicallymeshseveraldatasourcesinasinglevisualizationcomponent.Anotherrelevant
fact is that the API sometimes uses Javascript‐configurable Flash components: with this,
developers can combine the best of both AJAX and Flash in a single component. The
AnnotatedTimeLine (Figure15)enables thecreationof interactive timelineswith relative
ease.
Figure15‐Google'sAnnotatedTimeline
The Motion Chart (Figure 16) displays dynamic charts in real time. The component
draws both bar charts and point charts in real‐time. In addition, one can also add time‐
relateddatatothesechartsprovidingabetterunderstandingoftheevolutiononaspecific
case.Forinstance,onecanprovidedataaboutsalesresultsforagivencompanyonanentire
yearandpushthe“play”buttonandwatchresultsevolvinginthebrowser.
24
Figure16‐Google'sMotionChart
Aspreviouslymentioned,mapsare currentlyon topof thepreferences fordisplaying
geographical data [16]. This is mostly due to the fact that several interactivemaps have
appeared online. Google is no exception and provides a comprehensive API allowing the
creationofnovelapplicationsthatneedtodisplaygeographicalinformation.Itispossibleto
createseveralvisuallayersofinformation,mapzonesorevenreal‐timeevolvingmaps.
Despite the fact that there are lotsofmodern trends indisplayingdata, the simplest
format is still widely used: tables. The Google Visualization API also provides a Table API
where one can create interactive tables and present them to users. This concept goes
beyond staticHTML tables as thesenew tables canbe sorted andpaginated in real‐time.
Figure17showsadynamicGoogleTableinstance,containingthreedatafields.Thedatain
thetablecanbesortedbyanyofthefieldsand,ifrequired,canevenbeeditedinrealtime.
Figure17‐Google'sTable
25
3.3.4Summary
Nowadays, it is easy, toanyexperienceddeveloper, to createan informationvisualization
applicationthatsolveshisdatarepresentationproblem.WhetherusingFlash,Silverlightor
AJAX,developerscancreateinteractivevisualizationsthatwillenrichtheirworkand,atthe
same time, present data to users in a captivating interface and offering an effective user
experience.
Inordertounderstandthegenericcapabilitiesofthedescribedtools,Table1presentsa
short comparison of the evaluated tools. Five fieldswere evaluated in a 1 to 5 scale; the
fieldsrepresentthemainfeaturesofthetools.EaseofUseclassifieshoweasyitistocreate
a simple solution using the select tool. Learning Curve is used to classify the process of
creatingadvancedvisualizationapplicationswith theselected tools.Adaptabilityevaluates
howtheselectedtoolisadaptabletoaparticularissueinaspecificscientificarea.Number
ofToolsdescribesthenumberanddiversityoftheavailableframeworksintheselectedtool.
Futurerepresentstheprobabilityoffurtherdevelopmentsthatwillsignificantlyenhancethe
selectedtool.Finally,Overallclassifiestheoverallqualityofthetool.
Table1‐Toolcomparison
Flash Silverlight Javascript GoogleAPI
EaseofUse
LearningCurve
Adaptability
NumberofTools
Future
Overall
Legend:
1/VeryWeak/VeryHard
2/Weak/Hard
3/Moderate
4/Strong/Easy
5/VeryStrong/VeryEasy
26
4.Requirements
Oneoftheoutcomesofthisreportistoprovideasimpleguideforsomegoodpracticesand
requirementsforfuturedevelopersofwebdatavisualizationapplications.Theideaisnotto
detail user or technical requirements of any particular application but to provide a
comprehensive list of topics that should be addressed by anyweb developer in order to
createricherInternetapplicationsthatresultinimproveduserexperiences.
4.1UserRequirementsThefollowinguserrequirementsshouldbeaccomplishedinanywebapplication,especially
inthecurrentera,whereapplicationdesignandappearancehasgainedsignificantrelevance
[17]. However, considering that one is trying to present large datasets with a modern
visualizationtechnique,thesefeaturesbecomeevenmorecrucial.
Easeofuse:webapplicationsshouldbeeasilyusedforpeoplewithhighcomputer
skills but also for people with no background in computer science. End‐users
sometimes face complexapplications thatonlydifficult their comprehensionand
understandingofthepresentedinformation.
Accessibility: web applications rely on the browsing and traditional browsing
actions to execute any operation. These actions are typically not considered for
handicapped users. Web applications should be designed with readable color
themesandfonts,andshouldalsomakekeyboardshortcutsavailableforthemost
relevantactions.
Help: traditionally, web applications do not contain help or tutorial sections.
However, these are relevant because users are used to having them in their
desktops application.Making this kind of information available will improve the
applicationusability.Thiscanbeachievedthroughsmallhelpboxes,videotutorials
ofthemainfunctionalitiesoramanualexemplifyinghowtousetheapplication’s
keyfeatures.
Fast and responsive: the application should not be slow. This means that even
when some process requires a large amount of time to execute, the application
27
shouldpresentfeedbacktotheusers,showingthatthesystem“ismoving”.With
AJAXpatterns,itiseasiertocreatethisfluidexecutionmechanism,wheretheuser
realizeswhatishappeninginsidetheapplicationthroughaninterfacehintinstead
ofstaringatablankscreenwhilethepageisbeingloaded.Besideshavingafluid
interface,theapplicationshouldalsobefaster inthebackground:SteveSouders’
High PerformanceWeb Sites [18] is the best source forwhat aspects should be
takenintoaccountwhendevelopinganovelwebapplication.
4.2InterfaceThe user interface is themain contact point between the application and the users. It is
throughitthatusersinteractwiththeinformation,meaningthatitiswhattheuserswillsee
and use. Therefore, it is important for anyweb developer to follow a broad collection of
goodpracticesthatimprovetheoverallapplicationinterfacequality.Someofthequalities,
thatuserinterfacesshouldhave,inordertoenhanceusability[19],arethefollowing:
Clean:theuserinterfaceshouldbeascleanaspossible,reducingworkspaceclutter
and the amount of data shown to users. It is better to present less information
withmorequality thanmore information filling theentirewebpagewithoutany
visibledistinction.
Light:webapplicationsdependontheavailablebandwidthtooperateefficiently.If
the interface is clutteredwith heavy elements such as images and scripts it will
requirelongerloadingtimes,constrainingtheuserexperiencefromthestart.
Usable heuristics: a good idea to design human‐computer interfaces is to study
Nielsen’sUsabilityHeuristics[20].Theseheuristicshavegainedsomerelevancein
the human‐computer interaction world and provide a good setup for any
developerordesigner.However, theyarenotobliged to follow these references
buttheyareaninitialsteptoproducequalitywork.
28
5.Conclusion
The Internet has grown to provide themost complete application environment known to
date.With this growth, applicationshave changed. TheWebcannowbeunderstoodasa
completeapplicationplatform,overcomingmanyofthedesktopfaults.Withthisinmind,it
isgettingmoreandmoreimportantthatresearcherspublishtheirdiscoveriesandresearch
workonline.
Visualizationtechniquesare,sincealongtimeago,helpfultoanyresearcherthatneeds
to display large datasets of information in a comprehensive manner. Tables, charts and
manyother typesofgraphsare traditional techniques to facilitateusers’ taskofanalyzing
largeamountsofdata.Combiningboththesekeyideas,itisobviousthattheInternethasto
supportthesedatavisualizationtechniquesandenhancethem,creatinginterfacesthatwere
notpossibleifoneonlyusedcloseddesktopapplications.
Thisreportdetailssomeofthemoderntechniquesthatarereplacingstatictablesand
charts as visualization components. Additionally, it presents tools to implement these
techniques thatcanbeeasilyadoptedbyanyresearcher topresenthisdiscoveriesonline.
Interactive timelines, maps or motion charts are some of these techniques. They can be
easilydeployedwiththedescribedframeworks.TheseFlash,SilverlightorAJAXframeworks
will empower researchers with the tools needed to create next‐generation visualization
componentsandpresentlargedatasetsofinformation.Thesecomponentswillgiveusersa
faster insightontheknowledgeresearchersarepresenting,thus, increasingresultvisibility
andoverallusability.
29
References
1. Lopes,P., IntegraçãodeServiçosparaExtracçãodeConhecimento, inElectronics,Telecommunications and Informatics Department. 2008, University of Aveiro:Aveiro.
2. Rohrer, R.M. and E. Swing, Webbased information visualization. ComputerGraphicsandApplications,IEEE,1997.17(4):p.52‐59.
3. Chi, E.H., Improving Web usability through visualization. Internet Computing,IEEE,2002.6(2):p.64‐71.
4. Olivares R, J.C., et al., Using Web Pages Accessible Design for the Correct WebVisualization. Latin America Transactions, IEEE (Revista IEEE America Latina),2007.5(2):p.122‐127.
5. Eick,S.G.,etal.GeoBoost:AnAJAXWeb2.0CollaborativeGeospatialVisualizationFramework.inAerospaceConference,2007IEEE.2007.
6. Eibe, S. and O. Marban.WVFA: a Web visualization framework architecture. inIEEEInternationalConferenceonInformationReuseandIntegration.2004.
7. Wei,P.‐Y.ViolaHomePage.2008[cited20092009‐05‐15].8. Jern, M. 3D data visualization on the Web. in Conference Proceeding of the
MultimediaModelingConferecen,MMM'98.1998.9. Jianghui,Y.,D.Gracanin,andL.Chang‐Tien.Webvisualizationofgeospatialdata
using SVG and VRML/X3D. in Image and Graphics, 2004. Proceedings. ThirdInternationalConferenceon.2004.
10. Lai Feng,M.,A. Sourin, andK. Levinski.Functionbased3Dweb visualization. inCyberWorlds,2002.Proceedings.FirstInternationalSymposiumon.2002.
11. Chen,C.,Top10unsolvedinformationvisualizationproblems.ComputerGraphicsandApplications,IEEE,2005.25(4):p.12‐16.
12. Oreilly, T.,What is Web 2.0: Design Patterns and Business Models for the NextGenerationofSoftware.2007:SSRN.
13. Karam, G.M., Visualization using timelines, in Proceedings of the 1994 ACMSIGSOFT international symposium on Software testing and analysis. 1994, ACM:Seattle,Washington,UnitedStates.
14. Kumar, V., R. Furuta, andR.B. Allen,Metadata visualization for digital libraries:interactivetimelineeditingandreview,inProceedingsofthethirdACMconferenceonDigitallibraries.1998,ACM:Pittsburgh,Pennsylvania,UnitedStates.
15. Spangler, S., J.T. Kreulen, and J. Lessler.MindMap: utilizingmultiple taxonomiesand visualization to understand a document collection. inSystem Sciences, 2002.HICSS.Proceedingsofthe35thAnnualHawaiiInternationalConferenceon.2002.
16. Tufte, E.R. and G.M. Schmieg, The Visual Display of Quantitative Information.AmericanJournalofPhysics,1985.53(11):p.1117‐1118.
17. Wood, L.E.,User Interface Design: Bridging the Gap from User Requirements toDesign.1997:CRCPress,Inc.320.
18. Souders, S., High Performance Web Sites: Essential Knowledge for FrontEndEngineers:O'ReillyMedia,Inc.168.
19. Dumas, J.S.and J.C.Redish,APracticalGuidetoUsabilityTesting.1999: IntellectBooks.416.
30
20. Nielsen, J.Heuristics forUserInterfaceDesign. 20052005[cited20092009‐05‐13];Availablefrom:http://www.useit.com/papers/heuristic/heuristic_list.html.