Lecture Notes in Computer Science 6431 CommencedPublicationin1973 FoundingandFormerSeriesEditors: GerhardGoos,JurisHartmanis,andJanvanLeeuwen EditorialBoard DavidHutchison LancasterUniversity,UK TakeoKanade CarnegieMellonUniversity,Pittsburgh,PA,USA JosefKittler UniversityofSurrey,Guildford,UK JonM.Kleinberg CornellUniversity,Ithaca,NY,USA AlfredKobsa UniversityofCalifornia,Irvine,CA,USA FriedemannMattern ETHZurich,Switzerland JohnC.Mitchell StanfordUniversity,CA,USA MoniNaor WeizmannInstituteofScience,Rehovot,Israel OscarNierstrasz UniversityofBern,Switzerland C.PanduRangan IndianInstituteofTechnology,Madras,India BernhardSteffen TUDortmundUniversity,Germany MadhuSudan MicrosoftResearch,Cambridge,MA,USA DemetriTerzopoulos UniversityofCalifornia,LosAngeles,CA,USA DougTygar UniversityofCalifornia,Berkeley,CA,USA GerhardWeikum MaxPlanckInstituteforInformatics,Saarbruecken,Germany Achim Ebert Alan Dix Nahum D. Gershon Margit Pohl (Eds.) Human Aspects of Visualization Second IFIP WG 13.7 Workshop on Human-Computer Interaction andVisualization HCIV (INTERACT) 2009 Uppsala, Sweden,August 24, 2009 Revised Selected Papers 1 3 VolumeEditors AchimEbert UniversityofKaiserslautern,ComputerGraphicsandHCI Gottlieb-Daimler-Straße,67663Kaiserslautern,Germany E-mail:[email protected] AlanDix LancasterUniversity,ComputingDepartment Lancaster,LA14WA,UK E-mail:[email protected] NahumD.Gershon TheMITRECorporation 7515ColshireDrive,McLean,VA22102-7539,USA E-mail:[email protected] MargitPohl ViennaUniversityofTechnology InstituteofDesignandAssessmentofTechnology Favoritenstraße9-11/E187,1040Vienna,Austria E-mail:[email protected] ISSN0302-9743 e-ISSN1611-3349 ISBN978-3-642-19640-9 e-ISBN978-3-642-19641-6 DOI10.1007/978-3-642-19641-6 SpringerHeidelbergDordrechtLondonNewYork LibraryofCongressControlNumber:2011922135 CRSubjectClassification(1998):H.5,I.3-4,J.4 LNCSSublibrary:SL3–InformationSystemsandApplication,incl.Internet/Web andHCI ©Springer-VerlagBerlinHeidelberg2011 Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis concerned,specificallytherightsoftranslation,reprinting,re-useofillustrations,recitation,broadcasting, reproductiononmicrofilmsorinanyotherway,andstorageindatabanks.Duplicationofthispublication orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,1965, initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violationsareliable toprosecutionundertheGermanCopyrightLaw. Theuseofgeneraldescriptivenames,registerednames,trademarks,etc.inthispublicationdoesnotimply, evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevantprotectivelaws andregulationsandthereforefreeforgeneraluse. Typesetting:Camera-readybyauthor,dataconversionbyScientificPublishingServices,Chennai,India Printedonacid-freepaper SpringerispartofSpringerScience+BusinessMedia(www.springer.com) Preface Thedataavailabletousallisincreasingbecauseofthemyriadsensorsthatnow surround us and the networks and the Web, which links previously disparate data together. At the same time displays have changed, as well as the ubiqui- tous desktop or laptop screen we now also have tiny mobile phone displays as vast giga-pixeldisplay walls creating new challenges and opportunities for visu- alizingthisdataatbothendsofthescale.Innumerousareasincludingbusiness, healthcare, security, and environmental management the effective visualization andanalysisofdatais essentialto understandthe worldaroundandthe impact ofpotentialactions.However,thesevisualizationswillbe impairedifnotuseless unless they are based on a sound understanding of the way we as humans are able to comprehend and interact. This book is intended to give an overview of important issues concerning human–computer interaction and information visualization. It is based on an IFIP workshop that took place during the Interact 2009 conference in Uppsala. This workshop, co-organized by IFIP WG TC13.7 on “HCI and Visualization” and the European VisMaster Coordinated Action, surveyed and expanded our understanding of the cognitive and perceptual issues of interactive visualization and visual analytics and brought together researchersinterested in these issues. Ithighlightedtheresearchrequiredtounderstandwhataspectsofanalysismatch human capabilities most closely, and how interactive visual support should be designed and adapted to make optimal use of human capabilities in terms of information perception and processing. ThroughoutthediscussionsduringtheHumanAspectsofVisualizationWork- shop,variouscommontopicsemergedamongtheparticipantsandthesecommon topics are reflected in the papers in this volume.The papers indicate that there arestill manyopen researchquestionsevenwhen one considerstopicsthat have been studied for a considerable time such as fisheye interfaces. Furthermore, they remind us that visualization is no longer confined to expert users but may be used by everyone. The book can only outline some of the most prominent problems and some potential solutions to the effective application of human– computer interaction in information visualization. It does not provide a final answer to these issues, but offers a glimpse into an exciting area of growing importance to us all. December 2010 Achim Ebert Alan Dix Nahum Gershon Margit Pohl Table of Contents Human Aspects of Visualization Human Aspects of Visualization ................................... 1 Achim Ebert, Alan Dix, Nahum Gershon, and Margit Pohl The Broker...................................................... 10 Bob Spence Comparing Different Layouts of Tag Clouds: Findings on Visual Perception ...................................................... 23 Stephanie Deutsch, Johann Schrammel, and Manfred Tscheligi The PersonalEquation of Complex Individual Cognition during Visual Interface Interaction.............................................. 38 Tera Marie Green and Brian Fisher Faceted Visual Exploration of Semantic Data........................ 58 Philipp Heim and Ju¨rgen Ziegler Fisheye Interfaces—ResearchProblems and Practical Challenges ....... 76 Mikkel Rønne Jakobsen and Kasper Hornbæk Visualization of Workaday Data Clarified by Means of Wine Fingerprints..................................................... 92 Andreas Kerren Staying Focused: Highlighting-on-Demand as Situational Awareness Support for Groups in Multidisplay Environments.................... 108 Olga Kulyk, Tijs de Kler, Wim de Leeuw, Gerrit van der Veer, and Betsy van Dijk Using Gaze Data in Evaluating Interactive Visualizations ............. 127 Harri Siirtola and Kari-Jouko Ra¨iha¨ Giga-Scale Multiresolution Volume Rendering on Distributed Display Clusters ........................................................ 142 Sebastian Thelen, Joerg Meyer, Achim Ebert, and Hans Hagen Teaching Visual Design as a Holistic Enterprise...................... 163 Gerrit C. van der Veer and Corn´e Verbruggen Author Index.................................................. 173 Human Aspects of Visualization Achim Ebert1, Alan Dix2, Nahum Gershon3, and Margit Pohl4 1 Universityof Kaiserslautern, Kaiserslautern, Germany [email protected] 2 Lancaster University,Lancaster, UK [email protected] 3 The MITRE Corp., McLean, VA, USA [email protected] 4 ViennaUniversity of Technology, Vienna, Austria [email protected] 1 Introduction ThisbookisintendedtogiveanoverviewofimportantissuesconcerningHuman- Computer Interaction and information visualization. It is based on an IFIP [4] workshopthat took place during the Interact’09conference in Uppsala [3]. This workshop,co-organizedbyIFIPWGTC13.7on”HCIandVisualization”[2]and the European VisMaster Coordinated Action [5], surveyed and expanded our understanding of the cognitive and perceptual issues of interactive visualization and visual analytics and brought together researchersinterested in these issues. It outlined the research required to understand what aspects of analysis match human capabilities most closely, and how interactive visual support should be designed and adapted to make optimal use of human capabilities in terms of information perception and processing. Thisisachallengingagendathathastotakeintoaccountmanyfactorsinclud- ing user roles,tasks, collaborations,interests and previous knowledge as well as understanding the capabilities of display devices and computation methods for dealing with often very large data sets. The workshop goals required the coop- erationof scientists from different disciplines addressing aspects of visualization and analysis. This enabled a mapping between foundational theories in the dif- ferent researchareas to support collaboration among researchersin these fields. This included researchers and developers already working at the intersection of HCI,visualization,visualanalytics,andrelatedareassuchasdata mining,data management, perception and cognition. 2 Goals and Issues Displaysaffectthevisualinteractionofhumanwithcomputerdevicesmorethan anyotherhardwarecomponent.The displaydetermines,forexample,the physi- calsizeofthe viewport,the rangeofeffectivefield-of-viewandresolutionthatis possible, and the types of feasible interaction modalities. All other interactions A.Ebertetal.(Eds.):HCIV(INTERACT)2009,LNCS6431,pp.1–9,2011. (cid:2)c IFIPInternationalFederationforInformationProcessing2011 2 A. Ebert et al. must be placed in the context of the display. Thus in recent years, two trends have been emerging. First, visualization on large displays has become an increasingly important research topic. This is manifested in many applications in the use of wall-sized display devices. For example, large public screens enable a bigger audience to view the contents. Large displays for collaborative work provide the necessary screen real estate to allow the display of different points of view of the different users. The use of large display areas enables a more thorough immersion in virtual reality applications. And finally, large high-resolution displays enable visualization of large and complex datasets by facilitating both overview and detail views at the same time. On the other hand, the proliferation of mobile devices requires visual appli- cations on very small screens. With these smaller-scaledisplays, assumptions of thenormaldesktopmetaphornolongerapplyandmiceandkeyboards,designed for desktop interaction, become unsatisfactory. This calls for novel interaction devices and techniques designed for small screens of mobile devices. Regardless of the device in use, interactive visualization methods are needed to augment cognitive resources by providing an additional, external visual re- source to the human memory. They may reduce the amount of searching and make it easier to recognize patterns as well as enhance the understanding of relationships especially in large amounts of data and information. In addition, interactivevisualizationmethodsprovideamediumthatenablestheusertohave a representationofinformationthathe orshe canquicklyandeasily modify, re- structure or consider from a different perspective. This ability to manipulate the data is of extreme importance especially for analytical reasoning and sense- making. ResearchinHCIandvisualizationhasproducednumerousformsofvisualiza- tionandmethods forinteractingwithvisualizationsoflargedatasets.However, systematic understanding of the interaction issues with visualization and the human cognitive and perceptual processes involved is perhaps less well devel- oped. Therefore, the workshop at Interact 2009 tackled the human aspects of visualizationandbuiltaninspiringplatformforpresentationsandintensiveand controversial discussions. The articles in this book capture the exciting presen- tations and discussions about these issues. 3 The Articles in This Book The following section is a short overview of the papers in this book. All but the paper of Bob Spence – the keynote speaker at the workshop – are listed in alphabetical order (name of the first author). 3.1 The Broker Bob Spence (Keynote) There is a considerable amount of existing perceptual and cognitive knowledge that is highly relevant to information visualization. Nevertheless, it is difficult HumanAspects of Visualization 3 to directly apply it to the design of information visualization. Much of the re- searchis very abstract in nature. Some results might be context-dependent and could not be easily generalized to information visualization applications. In his keynote lecture, Bob Spence, therefore, advocated that ”Brokers” should me- diate between cognitive scientists and information visualizationdesigners.”The taskofaBrokeristointerpretrelevantknowledgeacquiredbycognitiveandper- ceptualpsychologistsandbringitsuitablytothenoticeofinteractiondesigners, thereby avoiding the need for that designer to have knowledge of cognition and perception.” A suitable method to do this brokerage is to develop Design Actions. Design Actions are structured forms that provide design guidance at a useful level of detail.Inaway,theytranslateknowledgefromcognitivepsychologyintoa”lan- guage” that can be understood by designers. Basically, they are guidelines, but they also contain information about the empirical evidence supporting them, and advantages and disadvantages of following the guideline. Bob Spence’s article provides a valuable foundation for the following papers, many of which try to connect psychologicalresearchwith design issues of infor- mation visualization. 3.2 Comparing Different Layouts of Tag Clouds: Findings on Visual Perception Stephanie Deutsch, Johann Schrammel, Manfred Tscheligi Tag clouds are an example of casual information visualization as described in the article of Andreas Kerren (see below). They can support users who want to getacomprehensiveoverviewoftheannotationsthatareoftenchaoticallymade bycommunitiesontheweb.Clusteringcanproducemeaningfulunitswithintag clouds,buttheseunitshavetobepresentedinawaythatcanbeeasilyperceived and understood. The authors argue that, while tag clouds are intuitively comprehensible and attractive for users, they have certainlimitations, e.g. the amount of redundant annotations or the problem that concepts are understood differently by various users.Inaddition,morepossibilitiesforinteractionwouldbeadvantageous.They also emphasize the importance of developing a solid foundation in perceptual psychology,especiallywithregardtovisualperceptionandeyemovement.Their empirical study indicates that semantic clustering (in contrast to traditional methods of tag-cloud-generation) might be useful for classifying annotations. However,they point outthat there are still openquestions regardingthe design of such visualizations. 3.3 The Personal Equation of Complex Individual Cognition During Visual Interface Interaction Tera Marie Green, Brian Fisher GreenandFisherdiscussapproachesfromcognitivepsychologythatmightguide the design of interactive visual interfaces. They point out that the interaction 4 A. Ebert et al. with visual interfaces consists of complex cognitive behavior. So far, evaluation studies have focused on low-levelcognitive processes (e.g. pre-attentive process- ing). This turns out to be interesting but not sufficient. In this context, they discuss theories of reasoning, especially the so-called sense-making loop. They point out that human reasoning processes can vary widely and therefore it is noteasyto developcomprehensivetheoriesofhumanreasoning.Models likethe sense-making loop may form a general framework, but cannot guide concrete design decisions. The authors also argue that human cognition is fundamentally influenced by individual differences (e.g. how humans categorize information, how they per- ceive visually or which problem-solving behaviors they adopt). Based on this assumption,they developedthe HumanCognitionModel,anoperationalframe- work to guide the design of collaborative interactive systems. 3.4 Faceted Visual Exploration of Semantic Data Philipp Heim, Ju¨rgen Ziegler HeimandZiegleraddressthe problemofsearchingandexploringsemanticdata, in particular whether faceted search might be a possible way to support such search processes in Semantic Web data. In faceted search, the search space is divided based on independent search dimensions that serve as filters. To sup- port users in their explorationof semantic data they developed a graphicaltool representing facets and search results as nodes in a visualization that exploits thegraph-basedstructureoflinkedsemanticdata.Thisgraphicalrepresentation can help users to develop a coherent cognitive model of the target domain. Heim and Ziegler argue that searching semantic data differs from using rela- tionaldatabasesandconventionalinformationretrieval.Thetooltheydeveloped is specifically targeted at searching and exploring semantic data and especially supports exploration and hypotheses refinement processes. 3.5 Fisheye Interfaces – Research Problems and Practical Challenges Mikkel Rønne Jakobsen, Kasper Hornbæk One of the major problems in information visualization is the representation of large amounts of data and information. Cognitive overload resulting from the display of large amounts of data has to be avoided by an appropriate and ef- fective design. There are several possibilities to reach this goal, one of which is focus+context interfaces (e.g. fisheye interfaces). Although using fisheye inter- facesisawell-knowntechniquedatingbackmanyyears,JakobsenandHornbæk arguein their article thatmany problems concerningthis technique arestill not solved. They point out that, for example, the notion of focus is not very well defined in the literature. Consequently, different systems are developed with a large variety of interaction strategies. This makes it difficult to give an overall