ix Preface noitamrofnI visualization has become a major theme during dimensional aspects, the users of information visualization the past ten years for much of the work of the faculty, tools are interested in finding relationships among variables; staff, and students at the University of Maryland's discovering similar items; and identifying patterns such as Human-Computer Interaction Lab (HCIL). Our roots are in clusters, outliers, and gaps. human-computer interaction (HCI), but increasingly our Another important discriminator is that scientific visual- attention has focused on information visualization. ehT reason ization users are primarily interested in continuous variables is clear: The overall theme of our work is to improve the such as density, temperature, or pressure, whereas information experience of people using computers, making that experience visualization users deal with continuous as well as categorical more effective and enjoyable. In order to reach that goal, we variables, such as gender, race, home ownership, date of birth, must create designs that enable users to develop control over state name, and number of bedrooms. Another distinctive the computer and, we hope, to attain a sense of mastery. For feature of information visualization is its attention to discrete this to occur, users must have a fluid and efficient interaction structures such as trees and graphs. Of course, there are areas with the computer--and the high bandwidth of visual where ideas and applications cross over, but the distinctive interfaces is a compelling way to attain this goal. aspects of information visualization are important to understand. A large component of human perception of the world si The interactive nature of information visualization stems through sight. As we've said before, eht eyes have .ti There si from the use of powerful widgets that enable users to simply more bandwidth and processing power rof input explore patterns, test hypotheses, discover exceptions, and through the human eyes than through any other sensory explain what they find to others. Interacting with the data modality. Sound, touch, smell, and taste are important, but the set gives users the chance to rapidly gain an overview, HCIL's researchers have repeatedly returned to highly visual explore subsets, or probe for extreme values. Information and dynamic displays as the best way to solve a surprisingly visualization tools become telescopes and microscopes that broad set of problems in an equally broad set of domains. allow users to see phenomena that were previously hidden. The HCIL is not alone in this belief. The field of information A steadily growing set of books si helping to define the visualization has been maturing, along with its related field and support graduate courses in many universities. The discipline of scientific visualization. Both have strong ties to classic book by Bertin, Semiology of scihparG (1983), has the broader field of human-computer interaction, as well as inspired many researchers, while the more recent sgnidaeR ni to graphics groups such as the ACM Special Interest Group Information Visualization: Using Vision ot Think (Card, on Graphics (SIGGRAPH) and many international graphics Mackinlay, and Shneiderman, )9991 has stimulated numerous associations. Information visualization has grown rapidly developers. The latter book includes 74 early papers from since ,5991 with annual conferences, organized in the United diverse sources with integrative commentaries and an extensive States by the IEEE and in the United Kingdom by the bibliography. Other books on visualization include the fine International Conference on Information Visualization. surveys by Robert Spence ,)0002( Colin Ware ,)0002( and An important distinction must be made between the Chaomei Chen .)9991( Conference proceedings are important more mature field of scientific visualization and the relative ,secruoser and snoitcelloc of papers on special scipot era common newcomer, information visualization. There are certainly in this discipline. Journals devoted to the topic such as overlaps, but scientific visualization researchers deal primarily Information Visualization (www.palgraveqournals.com/ivs/) with three-dimensional physical objects and processes such serve to present current research. Guides for practitioners are as blood flowing through heart valves, tornado formation, beginning to emerge (Westphal and Blaxton, .)9991 crystal growth, protein structures, and oil reservoir shapes. The broader field of HCI is now also firmly established They focus on volumes and surfaces, studying formations with major groups such as ACM's Special Interest Group on and flows and asking questions about inside/outside, Computer Human Interaction (SIGCHI), Usability above/below, or left/right. Professionals Association (UPA), British Computer Society yB contrast, information visualization researchers are Human Computer Interaction Group, and the Association concerned with abstract phenomena for which there may not Fran~ais pour l'Interaction Homme-Machine (AFIHM). be a natural physical reality, such as stock market movements, These organizations have significant participation not only social relationships, gene expression levels, manufacturing from academic researchers, but also from companies and production monitoring, survey data from political polls, or governmental organizations. In fact, the premiere HCI supermarket purchases. While both kinds of data sets come conference (SIGCHI) typically has attendance split equally from the physical world, instead of dealing with three- between researchers and practitioners. x Preface THANKS TO THE HCIL COMMUNITY to focus on information visualization, leaving our colleagues On this 20th anniversary of the HCIL, we proudly bring to publish other specialized books, such as the collections on together this collection of work from our colleagues-- children's technology by Allison Druin ,9991( .)0002 students, staff, visitors, and collaborators around the world. The work in this book could have been done only with the eW hope that by offering this work along with our reflections participation of a terrific staff. Research assistants, like Anne on what was important and why, and how the research Rose, have stuck with us through thick and thin, contributing unfolded as it did, we can shed some light on the often broadly to our research. Our newer staff, Aaron Clamage, mysterious process of innovation and creation--and encourage Allison Farber, Jesse Grosjean, Trina Harris, and Sabrina others to fftrther advance the field of information visualization. Liao, have already made their marks, and we are excited eW are deeply indebted to all of our colleagues at the about their joining us. University of Maryland and around the world. The rich The rhythm of our work is tied to the seasons--fall, spring, intellectual atmosphere and warm, personable climate in and summer semesters. A new semester is a chance to learn which we work has made this book possible. Our close faculty from the past, and each affords an opportunity to start fresh. colleagues, Fran~;ois Guimbreti6re in Computer Science, eW have had extremely strong computer science doctoral stu- Kent Norman in Psychology, Allison Druin, and Doug Oard dents who have gone on to make notable accomplishments of in Information Studies are true partners in the HCIL. Gary their own. Master's students and undergraduates have also Marchionini was an important participant for many years made important contributions and are co-authors on many before he moved to University of North Carolina. The papers. Our students, present and past, are what make the HCIL's long-term research scientist, Catherine Plaisant, has HCIL such a dynamic place. been a key developer of many of the ideas in this book. Other contributors have been yearlong visitors and postdoctoral Benjamin .B Bederson researchers such as Richard Biegel, Khoa Doan, Richard bederson @ .sc .dmu edu Salter, and Jean-Daniel Fekete. neT years ago, THE HCIL published a book called skrapS of neB namredienhS [email protected] noitavonnI ni retupmo.C-namuH noitcaretnI ,)3991( containing a selection of the lab's work from its first ten years. This decade, egelloC ,kraP Maryland on the other hand, has been so fruitful that we have decided ,yraunaJ 3002 xi stnemgdelwonkcA CHAPTER OPENER QUOTES Introduction Morris Kline, in Rosner, Stanley and Abt, Lawrence .E (eds.). ehT Creative .ecneirepxE :0791 Dell Publishers, p. .49 Rosner, Stanley and Abt, Lawrence .E (eds.) (ibid). p. .183 Chapter 1 John Tukey. "The Technical Tools of Statistics." American Statistician, 91 .)5691( Chapter 2 Leonardo da Vinci, .9151-2541 Chapter 3 Eliel Saarinen, as quoted by his son Eero, in Time, June ,2 .7791 Chapter 4 Ralph Waldo Emerson. Society dna Solitude .)0781( 2001: Fredonia Books. Chapter 5 Arthur Schopenhauer. agreraP dna .anemopilaraP .loV ,2 Ch. ,22 Sct. 752 .)1581( :1002 Oxford University Press. Chapter 6 Mickey Hart, Drumming at eht Edge of :cigaM A Journey into eht Spirit of .noissucreP :0991 Harper SanFrancisco. p. .99 Chapter 7 Don Norman, Things That Make sU Smart: Defending Human Attributes in eht Age of eht .enihcaM :3991 Perseus Publishing. Chapter 8 Jennifer Preece, as quoted to neB Shneiderman in a personal conversation, .1002 FIGURES ! N INTRODUCTORY MATERIAL 1.1 Shneiderman, ,.B Dynamic Queries for Visual Information Seeking. IEEE Software, vol. 11 )6( (Nov. 1994) 70-77, Fig. .1 Copyright (cid:14)9 ,4991 .EEEI Used with permission. 2.1 Ibid, Fig. .3 Copyright (cid:14)9 ,4991 IEEE. Used with permission. 3.1 Courtesy of University of Maryland. 4.1 Courtesy of Christopher Ahlberg, Spotfire. moc.eriYops.www 5.1 Courtesy of University of Maryland. 1.2 Copyright (cid:14)9 International Center of Photography 9ro.pci.www 2.3 Bederson, .B ,.B Hollan, .J ,.D Perlin, ,.K Meyer, ,.J Bacon, ,.D and Furnas, .G .W .)6991( Pad++: A Zoomable Graphical Sketchpad for Exploring Alternate Interface Physics. lanruoJ of Visual segaugnaL dna Computing, ,7 pp. 3-31. Copyright (cid:14)9 Elsevier, .6991 4.2a-d Plaisant, C. Guide to Opportunities in Volunteer Archaeology: Case Study of the Use of a Hypertext System in a Museum Exhibit. Berk .E and Devlin, .J (eds.), Hypertext/Hypermedia .koobdnaH :1991 McGraw-Hill. 4.4 Courtesy of Catherine Plaisant, Maya Venkatraman, Kawin Ngamkajomwiwat, Randy Barth, boB Harberts, and Wenlan Feng. 5.6 Fekete, J.-D. and Plaisant, .C Interactive Information Visualization of a Million Items. INFOVIS :2002 IEEE Symposium no Information ,noitazilausiV .2002 1.7 Courtesy of Catherine Plaisant and Anne Rose. 2.7 Courtesy of Catherine Plaisant and aiJ .iL PAPERS ,greblhA .C dna ,namredienhS .B .)4991( laustY noitamrofnI :gnikeeS Highway Incident Data. .corP pohskroW no weN smgidaraP ni thgiT gnilpuoC fo cimanyD Query sretliF with dleifratS .syalpsiD noitamrofnI noitazilausiV dna ,noitalupinaM .43-62 thgirypoC (cid:14)9 ,9991 MCA IHC 49' ecnerefnoC ,.corP .713-313 thgirypoC (cid:14)9 ,4991 noitaicossA rof Computing ,yrenihcaM .cnI detnirpeR with noitaicossA rof Computing ,yrenihcaM .cnI Reprinted with .noissimrep .noissimrep ,ninaT ,.E ,tnasialP ,.C and ,nanrredienhS .B .)0002( gninedaorB ,namredienhS .B .)4991( cimanyD Queries rof lausiV noitamrofnI sseccA ot egraL enilnO sesabataD by gnizilareneG Query .sweiverP .gnikeeS EEEI ,erawtfoS ,)6(11 .77-07 Copyright (cid:14)9 ,4991 .EEEI .corP muisopmyS no weN smgidaraP ni noitamrofnI noitazilausiV dna Used with .noissimrep ,MKIC--noitalupinaM .58-08 Copyright (cid:14)9 ,0002 noitaicossA rof gnitupmoC ,yrenihcaM .cnI Reprinted with .noissimrep ,noskirderF ,.A North, ,.C ,tnasialP ,.C and ,namredienhS .B .)9991( ,laropmeT lacihpargoeG and lacirogetaC snoitagerggA gnaD ,.G North ,.C and namredienhS .B .)1002( cimanyD seireuQ deweiV through Coordinated :syalpsiD A Case Study with and gnihsurB no Choropleth .spaM .corP lanoitanretnI ecnerefnoC xii Acknowledgments on Information Visualization ,1002 .467-757 Copyright (cid:14)9 ,1002 Viewing Digital Books Online. Forthcoming in Interacting With .EEEI Used with permission. ,sretupmoC .)6(41 Copyright (cid:14)9 ,3002 .reiveslE North, ,.C Shneiderman, ,.B and Plaisant, .C .)6991( User Nation, D.A., Plaisant, ,.C Marchionini, ,.G and Komlodi, .A Controlled Overviews of an Image Library: A Case Study fo eht .)7991( Visualizing Websites Using a Hierarchical elbaT fo elbisiV Human. .corP ts1 ACM International ecnerefnoC no latigiD Contents :resworB .COTbeW .corP dr3 ecnerefnoC no Human srotcaF ,seirarbiL .28--47 Copyright @ ,6991 Association rof Computing dna eht Web. Courtesy fo Catherine Plaisant. Machinery, Inc. Reprinted with permission. Kandogan, .E and Shneiderman, .B .)7991( Elastic Windows: A Shneiderman, .B and Kang, .H .)0002( Direct Annotation: A Drag- Hierarchical Multi-Window World-Wide beW Browser .corP ACM and-Drop Strategy for Labeling Photos. .corP lanoitanretnI UIST ,79 .771-961 Copyright @ ,7991 Association for Computing ecnerefnoC noitamrofnI noitasilausiV ,)0002W( .59-88 Copyright (cid:14)9 ,0002 ,yrenihcaM Inc. Reprinted with permission. .EEEI Used with permission. Hightower, ,.R Ring, ,.L Helfman, ,.J Bederson, .B ,.B and Hollan, Bederson, .B .B .)1002( PhotoMesa: A Zoomable Image Browser .J .)8991( Graphical Multiscale beW Histories: A Study of Using Quantum Treemaps and Bubblernaps. .corP ecnerefnoC no PadPrints. .corP ACM ecnerefnoC no Hypertext (Hypertext ,)89 resU ecafretnI dna erawtfoS ygolonhceT TSIU( ,)1002 .08-17 .56-85 Copyright (cid:14)9 ,8991 Association rof Computing ,yrenihcaM Copyright @ ,1002 Association rof Computing Machinery, .cnI .cnI Reprinted with permission. Reprinted with permission. Asahi, ,.T Turo, ,.D and Shneiderman, .B .)5991( Visual -noisiceD Shneiderman, ,.B Kang, H., Kules, ,.B Plaisant, ,.C Rose, ,.A and :gnikaM Using Treemaps rof the Analytic Hierarchy Process. CHI ,riehcuR .R .)2002( A Photo History of :IHCGIS Evolution fo 59 oediV ,margorP abstract ni ACM CHI 59 ecnerefnoC ,noinapmoC Design from Personal to Public. ACM ,snoitcaretnI ,)3(9 .32-71 .604-504 Copyright @ ,5991 Association rof Computing Copyright @ ,2002 Association rof Computing Machinery, .cnI ,yrenihcaM Inc. Reprinted with permission. Reprinted with permission. ,oruT .D .)4991( Hierarchical Visualization with Treemaps: Making Bederson, .B .B and Boltman, .A .)9991( Does Animation Help esneS fo Pro Basketball Data. CHI 49 Video ,margorP abstract ni Users Build Mental Maps of Spatial Information? .corP InfoViz ,99' ACM CHI 49 ecnerefnoC ,noinapmoC .244--144 Copyright @ ,4991 .53--82 Copyright @ ,9991 .EEEI Used with permission. Association rof Computing Machinery, Inc. Reprinted with Bederson, .B ,.B Meyer, ,.J and Good, .L .)0002( Jazz: An elbisnetxE permission. Zoomable User Interface Graphics tiKlooT in Java. .corP UIST ,ranruK ,.H Plaisant, ,.C Teittinen, ,.M and Shneiderman, .B .)4991( ,0002 .081-171 Copyright @ ,0002 Association rof Computing lausiV Information Management for Network Configuration. ,yrenihcaM Inc. Reprinted with permission. lacinhceT Report ,8823-RT-SC University of Maryland, Good, .L and Bederson, .B .B .)2002( Zoornable User Interfaces sa Department fo Computer .ecneicS Courtesy fo University fo a Medium rof Slide Show Presentations. noitamrofnI noitazilausiV Maryland. ,)1(1 35-49. Copyright (cid:14)9 ,2002 Palgrave Macmillan Ltd. Reprinted ,nosredeB .B ,.B Shneiderman, ,.B and Wattenberg, .M .)2002( with permission. Ordered and Quantum Treemaps: Making evitceffE Use of D2 ,ke~bmoH ,.K Bederson, .B .B and Plaisant, .C .)2002( Navigation Space to Display Hierarchies. ACM snoitcasnarT no ,scihparG ,)4(12 Patterns and Usability of Zoomable User Interfaces with and .458-338 Copyright (cid:14)9 ,2002 Association for Computing without an Overview. ACM snoitcasnarT no Computer-Human ,yrenihcaM Inc. Reprinted with permission. ,noitcaretnI ,)4(9 .893-263 Copyright @ ,2002 Association rof ,etekeF .D-.J and Plaisant, .C .)2002( Interactive Information Computing Machinery, Inc. Reprinted with permission. Visualization of a Million Items. .corP IEEE InfoVis 2002,117-124. Plaisant, ,.C Marchionini, ,.G Bruns, ,.T Kornlodi, ,.A and Copyright (cid:14)9 2002 .EEEI Used with permission. Campbell, .L .)7991( Bringing Treasures to the Surface: Iterative Plaisant, ,.C Grosjean, ,.J and Bederson, .B .B .)2002( :eerTecapS Design for the Library of Congress National Digital Library Supporting Exploration in Large Node Link ,eerT Design Program. .corP CH197, .525-815 Copyright @ ,7991 Association rof Evolution and Empirical Evaluation. .corP IEEE InfoVis ,2002 Computing Machinery, Inc. Reprinted with permission. .46--75 Copyright @ ,2002 .EEEI Used with permission. ,esoR ,.A Ding, ,.W ,ininoihcraM ,.G elaeB ,.rJ ,.J and Nolet, .V .)8991( ,nosredeB .B .B .)0002( Fisheye Menus. .corP UIST 2000, .522-712 Building an Electronic Learning Community: From Design ot Copyright @ ,0002 Association rof Computing Machinery, .cnI .noitatnemelpmI .corP ,891HC .012-302 Copyright (cid:14)9 ,8991 noitaicossA Reprinted with permission. rof Computing ;yrenihcaM .cnI Reprinted with .noissimrep Plaisant, ,.C Mushlin, ,.R Snyder, ,.A ,iL ,.J Heller, ,.D and Shneiderman, ,.B Feldman, ,.D Rose, ,.A and Ferr6 Grau, .X .)0002( Shneiderman, .B .)8991( :seniLefiL Using Visualization to Enhance Visualizing Digital Library Search Results with Categorical and Navigation and Analysis fo Patient Records. 1998 American Hierarchial Axes. .corP 5th ACM International ecnerefnoC no latigiD lacideM scitamrofnI noitaicossA Annual Fail Symposium, .08-67 ,seirarbiL 57-66. Copyright (cid:14)9 ,0002 Association for Computing Copyright @ ,8991 American Medical Informatics .noitaicossA ,yrenihcaM Inc. Reprinted with permission. Used with permission. Druin ,.A Bederson .B ,.B Hourcade .J ,.P Sherman ,.L Revelle ,.G Hochheiser, .H and Shneiderman, .B ,)1002( Interactive Platner ,.M and Weng .S .)1002( Designing a Digital Library rof Exploration fo em53 Series Data. .corP Discovery ecneicS ht4 gnuoY Children: An Intergenerational Partnership. .corP lanoitanretnI ecnerefnoC 2001, .644-144 Copyright (cid:14)9 1002 Springer- ACMflEEE Joint ecnerefnoC no Digital ,seirarbiL .504-893 Copyright .galreV Used with permission. @ ,1002 Association for Computing Machinery, Inc. Reprinted ,etekeF .D-.J and Plaisant, .C .)9991( Excentric Labeling: Dynamic with permission. Neighborhood Labeling for Data Visualization. sgnideecorP of CHI Hourcade, ,.J Bederson, .B ,.B Druin, ,.A Rose, ,.A Farber, ,.A and ,99 .915-215 Copyright (cid:14)9 ,9991 Association for Computing ,anrayakaT .Y ,)2002( ehT International Children's Digital :yrarbiL ,yrenihcaM Inc. Reprinted with permission. stnemgdelwonkcA xiii ,nosredeB .B .B Czerwinski, ,.M and ,nostreboR .G .A .)2002( Shneiderman, .B .)6991( ehT Eyes Have :tI A ksaT by Data epyT Fisheye Calendar Interface rof :sADP Providing Overviews rof ymonoxaT rof noitamrofnI .snoitazilausiV .corP 6991 EEEI ecnerefnoC Small Displays. lacinhceT Report ,8634-RT-SC University fo no Visual ,segaugnaL .343-633 Copyright (cid:14)9 ,6991 .EEEI Used with Maryland, Department fo Computer .ecneicS Courtesy fo .noissimrep University of Maryland. Shneiderman, .B .)1002( Supporting Creativity with Advanced ,oeS .J and Shneiderman, .B .)2002( Interactively gnirolpxE Information-Abundant User Interfaces. In Eamshaw, ,.R ,jdeuG Hierarchical Clustering Results. IEEE ,retupmoC ,)7(53 .68-08 ,.R naV Dam, ,.A and ,ecniV .J (eds.), dertneC--namuH ,gnitupmoC Copyright (cid:14)9 ,2002 .EEEI Used with .noissimrep Online Communities, dna Virtual Environments, galreV-regnirpS North, .C and Shneiderman, .B .)0002( rehtegoT-panS London, .084-964 Copyright @ ,1002 .galreV-regnirpS Reprinted :noitazilausiV A User Interface rof Coordinating snoitazilausiV aiv with .noissimrep Relational Schemata. .corP Advanced lausiV secafretnI 2000, .531--821 Shneiderman, .B .)2002( Inventing Discovery :slooT Combining Copyright (cid:14)9 ,0002 noitaicossA rof Computing ,yrenihcaM .cnI Information Visualization with Data Mining. noitamrofnI Reprinted with .noissimrep ,noitazilausiV ,)1(1 5-12. Reprinted from .corP yrevocsiD ecneicS ht4 Plaisant, ,.C Carr, ,.D and Shneiderman, .B .)5991( resworB-egamI lanoitanretnI ecnerefnoC 2001 erutceL( Notes in Computer ,ecneicS .loV ymonoxaT and Guidelines rof .srengiseD IEEE ,erawtfoS ,)2(,21 ,6222 galreV-regnirpS Heidelberg). Copyright (cid:14)9 ,1002 -regnirpS .23-12 Copyright (cid:14)9 ,5991 .EEEI Used with permission. .galreV Reprinted with permission. XV Introduction Visual images or visual understanding of what one is trying to do are definitely helpful... I would say understanding is achieved and results come more readily if one has a picture rather than by looking at a lot of formulas. Morris ,enilK ehT evitaerC ecneirepxE )0791( Our goals in writing and editing this book were to give a sense of control and confidence while making progress researchers and students an understanding of how ideas in toward their goals. information visualization evolve and spread. ehT dissemination When people experience this kind of focus, they sometimes of ideas si a fascinating and instructive process, especially if it refer to "being in the flow." Some psychologists refer to this involves your original ideas. It is a thrill to see someone adopt as optimal experience, a shorthand that describes the best your ideas or software, often refining them substantially as experience that one can hope *.rof And though it may first seem they apply them to some novel domain. Fortunately, we far afield from computer work in information visualization, as have had very positive collaborative experiences. eW have researchers let us consider it our ideal: to create computer repeatedly found that by being open with our own ideas and interfaces that enable users to forget they are using a computer honest about the source of others', a great harmony resultsm and think only of the important work they are accomplishing. with innovation, excitement, and rapid improvement. This book is about that process in innovation during the last ten of the lab's twenty years as we concentrated on the THE IMPORTANCE OF FLOW field of information visualization, a subfield of HCI. Another important theme that has pervaded our work is something that has come to be called flow (Mihaly EVALUATING OUR WORK Csikszentmihalyi, 1990). Over the years we have developed How do we assess our progress? Are we any nearer to our an intuition about what makes information visualization goal of creating interfaces that support flow? Tough internal (and other) interfaces work well, and we have discovered assessments---critiquing each other, challenging assumptions, congruence between these ideas and the concept of flow, an and demanding evidence help prevent us from falling into idea from the psychology literature. Though we don't have traps of wishful thinking. a strict formula for a successful interface, we know that a External reviews from colleagues add to our continuing few basic approaches do help. In general, we believe it is assessments. eW send drafts of papers to colleagues, invite important that the users stay in control and that the computer visitors to see our work in progress, and engage potential offers choices with appropriate feedback for user actions. users to try our software. eW appreciate good feedback, taking Conversely, computer-controlled interactions often lead to to heart constructive comments that push us to refine our work. unpredictable, and therefore unacceptable, interfaces. The next level of assessment comes from anonymous eW also have learned that people are primarily interested reviewers of conference papers, journal articles, and grant in focusing on their tasks and not on operating the interface proposals. eW discuss rejections and try to learn from them. ---and yet so much of a user's experience with a computer is Even when we disagree with reviewers, we try to examine manipulating widgets, resizing windows, and selecting how we might have told the story more effectively. Some of from menus. It is crucial that computers give users prompt our strongest papers have been shaped by tough reviewing and informative feedback at every step along the way. processes. ,yllaniF users must stay engaged in the task for their experience Published papers are the clearest signs of our progress-- to be effective in the long run. This means that the interface they have been validated by peer review, and they are must not be too complex or confusing as to alienate users, publicly available. Several members of our group appear nor so simplistic or condescending as to make them bored. high on the list of authors ordered by frequency of publication A computer interface that strikes the right balance can in HCI papers, conference presentations, and books enable users to concentrate on the task at hand. The computer .bibich.~a~( .)lmth.srohtua/gro becomes a "tool" in the best sense of the word--an extension Another imperfect but useful metric of success is of the user's body. Time passes quickly, and the users develop the number of references to a paper. The NEC Citeseer enO* ralupop koob dellac wolF yb Mihaly iyIahimtnezskisC )0991( sezirammus eht ydob of krow ni gnidnatsrednu namuh lamitpo .ecneirepxe xvi Introduction (citeseer.nj.nec.com/directory.html) index has a special section North. eW have also enjoyed long-running interactions with on human-computer interaction, and we are proud that the Clare-Marie Karat and John Karat at IBM. Special mention most cited paper for many years has been one of our works on goes to Nahum Gershon of Mitre, who has been an effective information visualization. Similarly, when PARC researchers champion and organizer for information visualization studied reference patterns in information visualization, they conferences and journals. found that their group's papers were cited most frequentl~ but University colleagues include Steven Roth at Carnegie- our lab came in second. Mellon University, Steve Feiner at Columbia University, Citations in academic papers are one manifestation of George Fumas at University of Michigan, John Stasko and our influence, but downloading our software is also quite Jim Foley at Georgia Tech, Andries Van Dam at Brown validating. More than 30,000 individuals have downloaded University, Jim Hollan now at UCSD, Saul Greenberg at PhotoMesa, one of our image browsers (Chapter .)2 University of Calgary, Robert Spence at Imperial College, Our software also influences commercial and government Keith Andrew at the University of Graz, and Alfred applications. Many of our ideas have become part of larger Inselberg at leT Aviv University. Another special mention success stories, such as SmartMoney's MarketMap goes to Edward Tufte at Yale University, who si well known for ,)pamtekram/moc.yenomtrams.~ww( the Hive Group's treemaps his independently published books ,3891( 1990, )7991 and for )moc.puorgevih.~ww( (Chapter ,)6 and Spotfire's -tops.~ww( his annual public lecture tourmwe regularly pay for students )nroc.erif visualization tools (Chapter .)1 Contributions to to attend when he swings through the Washington, D.C. area. important national and international projects include the There are many others, but these people form the core of Visible Human Explorer for the U.S. National Library of our community. eW jointly write books and articles, organize Medicine, NASA's Earth Science Information Partnerships, conferences, and participate in workshops--all to promote and the Library of Congress's American Memory beW site. information ~;isualization to broader circles. Seeing each ;yllaniF an important internal measure of the HCIUs success other for a beer or dinner once a year is important, and the is the frequency with which our students graduate and join continuity of contact is maintained by email. eW tell our latest companies, universities, or government agencies where they stories, probe for their new ideas, and seek each other's make valuable contributions. It's especially satisfying to see respect. These colleagues are who we turn to validate our young, often shy or quiet students become self-confident innovations, to ask for reviews of our draft papers, and to be professionals who are valued by employers and respected by our partners in proposals. colleagues. As a community, we are gratified when former students return to tell us how much their time at the the HCIL THE MARYLAND WAY influenced them both professionally and personally. FOR INFORMATION VISUALIZATION WORKING WITHIN A BROADER ... feelings of excitement and pleasure accompany evitaerc COMMUNITY OF SCIENTISTS work. In many instances, this excitement si detaicossa with It si difficult ot rank or even list lla the people in our professional arribing at insights, seeing new principles, and discovering networks, so we must begin with an apology to anyone we spihsnoitaler which were not fully .detcepxe have left out in this discussion. Those who want completeness can examine the hundreds of references in our papers. yelnatS rensoR dna ecnerwaL .E Abt, However, we cannot honestly review our work without ehT Creative Experience I( 970) reflecting on the influences of our colleagues. In several ,snoitces we include more details related to that topic, but this opening Occasionally, visitors and colleagues who appreciate our mentions a few of the major groups whose influence cuts accomplishments will ask how we go about our work. In across many of the sections. skrapS of noitavonnI (1993), Ben Shneiderman described what Our strongest and most enduring bonds have been with he called eht dnalyraM Way. It has remained a useful guide. Of the community of researchers at the Xerox Palo Alto course, we've learned some new lessons, the field of HCI has Research Center, now simply PARC .)moc.crap.www( Our matured, and the lab has grown. oS we'd like to revisit those contacts have been mostly with the user interface group and ideas in the narrower context of information visualization. its related teams, especially the manager and long-term eW begin by choosing motivated, strong researchers who researcher, Stu Card. Stu is a leader in theory-driven thinking will interact well with others. Then, the Maryland Way is to and research and is a remarkable innovator, as testified by foster innovation through these seven steps. his numerous patents and papers. The PARC group has also .1 Choose a good driving problem. included key people such as Jock Mackinlay, George .2 Become immersed in related work. Robertson (now at Microsoft Research), Ramana Rao (now .3 Clarify short-term and long-term goals. at Inxight), Peter Pirolli, Mark Stefik, and many others. .4 Balance individual and group interests. As Microsoft Research grew, we enlarged our contacts with George Robertson, Mary Czerwinski, and others. Other .5 Work hard. industry groups include those at lleB-T&TA Labs and spinoff .6 Communicate with internal and external stakeholders. groups, including Stephen Eick, Andreas Buja, and Stephen .7 Get past failures. Celebrate success! Introduction xvii .1 Choose a Good Driving Problem often prefer chatting with the students who are doing the Fred Brooks's advice to choose a good driving problem is work to attending a private presentation by senior staff. especially relevant for information visualization, given its When visitors are potential funders, direct student and staff strong practical component. It helps enormously to have a involvement in the future of research projects increases clear goal; for example, design a video library that consumers motivation, participation, and work quality. can browse, or build a photo library program that three- When individual and group goals are in harmony, fortuitous year-olds can navigate. Finding good problems is like collaborations are likely. One of the ways we have been able to antique hunting: you are not quite sure what you want, but accomplish so much with limited resources si that individuals when you see it, you know it. In the early stages of choosing help each other. When one Ph.D. student needed a special a problem, we brainstorm to come up with alternatives. routine on an unfamiliar hardware and software environment, Then, over a period of a few weeks, we discard the extreme another student stepped in and provided a few days of ideas, refine the remaining possibilities, and focus on one. programming help. The favor was repaid by help in reviewing Our favorite problems entail improving designs for a wide paper drafts and in preparing subjects for an experiment. range of users in real-world contexts--building interfaces Since our lab operates with a diverse hardware and software for museum users exploring historical topics; library patrons environment, hardly an hour goes by without someone calling searching for a book or document; scientific researchers trying out for help on some system. to understand gene expression levels; or business analysts .5 Work Hard seeking patterns in customer behavior. Thomas Edison remarked that innovation requires I percent .2 Become Immersed in Related Work inspiration and 99 percent perspiration. An exciting and We expect each of our students to become the "world's leading novel idea si just the starting point. Most ideas have a cascade expert" on the problem he or she is investigating. Our students of smaller ideas behind them and details to be worked out. must acquaint themselves with related studies, sample similar Special cases, exceptions, and extreme conditions must be be commercial products, and personally contact active researchers. investigated carefully to reveal the limitations of a new idea. We expect our students to educate us about related work. Then converting an idea to a piece of functioning software, Our students must go to the library or the Intemet and a set of screen designs of a prototype, or the materials, tasks, chase down every reference to their topic. This process has and statistics in an experiment takes devoted effort. the dual benefits of compelling them to work on something Polishing, refining, and cleaning up can take ten or a hundred narrow enough that they can become the leading experts times more effort than the original innovation. Simply and forcing them to clarify exactly what they are working on. expecting things to take a great deal of effort removes some Trying out commercial products brings a sense of practical of the anxiety or expectation of perfection. reality. The students come to understand the parts in the There is a definite improvement in quality when you can context of the whole and to see the tradeoffs that designers revise a project after reflection or comments from colleagues. must make. The second time through almost any process or path is often Getting in touch with current researchers or developers is a smoother and fasten The "third-time charm" for experiments novel and threatening task for many students. Email helps or designs suggests that persistence and repeated tries leads facilitate the process, but phone calls, letters, and visits are also to excellence. important. Shy students overcome their awkwardness and are .6 Communicate with Internal and External Stakeholders often rewarded by a helping hand from a respected researcher Our group operates with a high degree of internal communication or an invitation to present their work at a major company. and external reporting. Internally, research teams working on .3 Clarify Short-Term and mreT-gnoL Goals related topics meet frequently. eW also hold weekly seminars to After the brainstorming process (see Step ,)1 which sharpens discuss journal or conference papers or to hear formal presenta- our understanding of the project, we establish long- and tions of results. Even more compelling than these traditional short-term goals. Long-term goals provide a destination and meetings, however, are the spontaneous demos, informal pre- a shared set of expectations that focuses effort. Short-term experiment reviews, partidpation in pilot studies, pleas for help goals provide immediate feedback about progress and a with statistics, and personal requests for reading draft papers. chance to make inexpensive midcourse corrections. While internal communication helps form and guide our .4 Balance Individual and Group Interests work, external communications increase our intensity as we We give each student or staff person a clear role that serves prepare for demonstrations to visitors; presentations at our his or her individual goals (e.g., getting a master's degree annual Symposium & Open House; writing reports, theses, within 81 months, doing an independent study summer project, or journal articles; production of videotape reports; lectures at or building a resume to get a desired job). Individual goals companies or universities; and papers and sessions at .secnerefnoc need to be in harmony with the overall goals and directions Preparing a presentation for a friend, staff person, or of the lab. A student who wants to do a master's thesis on a professor may encourage some diligent effort, but it seems topic that is poorly related to our existing work will be that preparation for a conference talk, a lecture to supporting encouraged to consider alternative topics. companies, or important visitors raises the stakes considerably. When visitors tour the lab, students and staff show their Telling the story and listening for feedback are often unfamiliar work, get feedback, and promote their ideas. Our visitors skills to technically oriented people, so we try to practice often. xviii Introduction .7 Get Past Failures. Celebrate Successes! and faculty dine with the advisory board to reflect on the Many days seem filled wihtuhn dreds responsibilities of such sa day and seek suggestions for future work. reviewing journalp apers, showing around, visitors responding In addition to the symposium, the HCIL holds an annual to requests reports, technical for writing or proposals, reading a all-day retreat at some bucolic location within an hour's draft of a thesis chapter. eW are sometimes burdened with drive of the bal (Figure .)2.0 eW discuss our directions, research filling out travel vouchers, repairing computers, or preparing envision the big picture, and make resolutions in a safe and budgets; however, when it comes time to write annual supportive environment. There si never enough time discuss to reports or prepare for our Symposium & Open House, we every project, but long lists are made rof later contemplation. are struck by how much we have accomplished during the CONCLUSION AND FUTURE DIRECTIONS previous year. The good days are when students invite us to see a Through these processes the HCIL has continued for over 02 demonstration of their latest design, improvement, or years to focus on topics that put humans at the center tech- of experiment. As lab members gather around a computer ;yalpsid nology. This book tells the story of the HCIL's work in infor- we cheer, comment, and criticize. Other memorable days mation visualization over the past decade. selected The papers include working intensely to finish a paper, resolving a which exhibit our lab's most important outcomes, show our problem with statistics, brainstorming on designs, rehearsing work process and ideas of evolution from one tcejorp ot the .txen rof videotape, a fantasizing futuusreer ,secafretni and yllaicepse Each chapter starts with our reflections on the people and celebrating a student's successful dissertaiton defense or problems that inspired the work, followed by the papers in journal submission. chronological order (except Chapter .)7 We've also included Even a successful research group must deal with short lists of favorite papers frooumt side our lab that were the disappointments. After many years of writing, it si still most relevant and influential. The listing of lla technical disappointing to be turned down by a conference program reports published by the HCIL in the last 01 years si presented committee or journal editorial board. Requests from journal ni( reverse chronological order) in Appendix D editors for major revisions are also unpleasant, but some of As a research topic, the field of information visualization our most successful papers have had the longest gestation si still forming with a growing number of university courses, periods and endured the most revisions. eW have had our professional conferences, and scientific journals. The central share of rejected grant applications, students who choose to research problems include perceptual psychology issues go elsewhere, and funders who decide not to renew. such as understanding change blindness, choosing color However, careful acknowledgement of contributors, palettes, showing relationships between nonproximal items, reviewers, and supporters of all kinds helps to keep such and using retinal properties size, (color, shape, etc.) properly. disappointing events to a minimum. In addition, we avoid Interface design research topics build on perceptual issues much internal strife by discussing author credits early and rof presenting information, providing user widgets, control and often and seeking creative ways to resolve conflicts. using animation .ylevitceffe There si also a need for traditional The HCIL's annual Symposium & Open House (Figure )1.0 computer science topics such as algorithms for rapid search, is a major celebration in which students, staff, and faculty data structures for compact storage, software architectures present their work to several hundred attending. In the implementation, efficient for anmdo dular programs to etatilicaf morning, we make formal presentations and respond to collaborative development. New research methodologies questions. The aftemoon si demonstrations, tours, to over given are needed to improve user-needs assessments, controlled and personal discussions. At the end of the day, senior staff experimentation, and ethnographic observations. erugiF 1.0 Picture of eht HC/L ekac from Open annual .esuoH Figure 2.0 HCIL srebmem at eht annual .taerter Introduction xix Central problems commercial for developers fo information ,draC ,.S ,yalnikcaM ,.J ,namredienhS .B ,).sde( sgnidaeR ni visualization tools include data integration to smoothly noitamrofnI :noitazilausiV Using noisiV ot Think, Morgan nnamfuaK ,srehsilbuP naS ocsicnarF .)9991( import data, data cleansing to remove or repair bad inputs, and data export to send result sets to other users in formats ,nehC ,iemoahC noitamrofnI ,noitazilausiV Springer galreV .)9991( that will be acceptable. Once these basic problems are ,iylahimtnezskisC ,ylahiM :wolF ehT ygolohcysP of lamitpO solved, commercial developers will succeed in crossing eht ,ecneirepxE ,snilloCrepraH weN kroY .)0991( chasm (to use Geoffrey Moore's term) fi they can provide a Drain, nosillA ,).de( ehT ngiseD of s'nerdlihC ,ygolonhceT nagroM whole product solution for a genuine need. nnamfuaK ,srehsilbuP naS ocsicnarF .)9991( When developers solve problems for their customers, ,niurD ,nosillA Hendler, semaJ ,).sde( stoboR for :sdiK gnirolpxE information visualization products will move from "nice-to- weN seigolonhceT for ,gninraeL Morgan nnamfuaK ,srehsilbuP naS have" to "must-have." Industries likely to be the early adopters ocsicnarF .)0002( are those driven by continuous innovation and repeated ,yeloF ,semaJ van ,maD Andries, ,renieF ,nevetS Hughes, ,nhoJ discovery, including pharmaceutical drug research, oil-gas retupmoC ,scihparG selpicnirP dna .,ecitcarP dn2 ed., -nosiddA exploration, financial analysis, and manufacturing quality ,yelseW ,gnidaeR .ssaM .)0991( control. Other candidate adopters are transportation safety ,ininoihcraM ,yraG noitamrofnI gnikeeS ni cinortcelE ,stnemnorivnE analysts, business fraud detectors, crime or terror investigators, egdirbmaC Cambridge Press, University .)5991( and medical diagnosticians. ,namredienhS .B ,).de( skrapS of noitavonnI ni retupmoc-namuh Researchers and product developers will have to cooper- ,noitcaretnI xelbA ,srehsilbuP Norwood, .J.N .)3991( ate in a massive educational process to teach potential users ,ecnepS ,treboR noitamrofnI ,noitazilausiV ,yelseW-nosiddA ,xessE about suitable applications and appropriate visualizations. dnalgnE .)1002( This process may take decades, just as it did for the move to ,etfuT Edward, ehT lausiV yalpsiD of evitatitnauQ ,noitamrofnI graphical user interfaces. Collaboration with data mining scihparG ,sserP ,erihsehC .nnoC .)3891( enthusiasts, statisticians, information technology specialists, ,etfuT Edward, gninoisivnE ,noitamrofnI scihparG ,sserP ,erihsehC software engineers, business analysts, and other profession- .nnoC .)0991( als will accelerate this adoption process. ,etfuT Edward, lausiV dna segamI :snoitanalpxE ,seititnauQ ecnedivE dna ,evitarraN Press, Graphics ,erihsehC .nnoC .)7991( ,eraW ,niloC noitamrofnI ,noitazilausiV nagroM nnamfuaK ,nitreB ,seuqcaJ ygoloimeS of ,scihparG University fo nisnocsiW ,sserP ,nosidaM .siW .)3891( ,srehsilbuP naS ocsicnarF .)0002(