reaching broader audiences with data visualization · visualization, data visualization on mobile...

9
Department: Visualization Viewpoints Editor: Theresa-Marie Rhyne, [email protected] Reaching Broader Audiences with Data Visualization Bongshin Lee Microsoft Research Eun Kyoung Choe University of Maryland, College Park Petra Isenberg Inria Kim Marriott Monash University John Stasko Georgia Institute of Technology Abstract—The visualization research community can and should reach broader audiences beyond data-savvy groups of people, because these audiences could also greatly benefit from visual access to data. In this paper, we discuss four research topics—personal data visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that, individually and collaboratively, would help us reach broader audiences with data visualization, making data more accessible. VISUALIZATION research has become a bur- geoning field encompassing many established and emerging sub areas, such as biological data vi- sualization, visualization for cyber security, and visual analytics for healthcare, among others. Such research progression has empowered ex- perts dealing with massive data (e.g., researchers, biologists, data analysts) to apply and leverage visualizations in their work. However, current visualization research often leaves out broader audiences, who could also benefit from innova- tions in visualization research. Compared to the experts mentioned above, these wider audiences have fundamental differences in their needs, mo- tivations, and goals as well as in their capabilities. We as a visualization research community should attune to these broader audiences as we see great opportunities to make practical impacts in their everyday life leveraging data visualization. In this paper, we discuss four research topics that can help us reach broader audiences with data visualization, making data more accessible. Ever growing capabilities of personal data col- lections provide a great opportunity for people Published by the IEEE Computer Society c 2020 IEEE 1

Upload: others

Post on 25-Sep-2020

36 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

Department: Visualization ViewpointsEditor: Theresa-Marie Rhyne, [email protected]

Reaching Broader Audienceswith Data VisualizationBongshin LeeMicrosoft Research

Eun Kyoung ChoeUniversity of Maryland, College Park

Petra IsenbergInria

Kim MarriottMonash University

John StaskoGeorgia Institute of Technology

Abstract—The visualization research community can and should reach broader audiencesbeyond data-savvy groups of people, because these audiences could also greatly benefit fromvisual access to data. In this paper, we discuss four research topics—personal datavisualization, data visualization on mobile devices, inclusive data visualization, and multimodalinteraction for data visualization—that, individually and collaboratively, would help us reachbroader audiences with data visualization, making data more accessible.

VISUALIZATION research has become a bur-geoning field encompassing many established andemerging sub areas, such as biological data vi-sualization, visualization for cyber security, andvisual analytics for healthcare, among others.Such research progression has empowered ex-perts dealing with massive data (e.g., researchers,biologists, data analysts) to apply and leveragevisualizations in their work. However, currentvisualization research often leaves out broaderaudiences, who could also benefit from innova-tions in visualization research. Compared to the

experts mentioned above, these wider audienceshave fundamental differences in their needs, mo-tivations, and goals as well as in their capabilities.We as a visualization research community shouldattune to these broader audiences as we see greatopportunities to make practical impacts in theireveryday life leveraging data visualization.

In this paper, we discuss four research topicsthat can help us reach broader audiences withdata visualization, making data more accessible.Ever growing capabilities of personal data col-lections provide a great opportunity for people

Published by the IEEE Computer Society c© 2020 IEEE 1

Page 2: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

Visualization Viewpoints

to make positive changes in their lives throughbetter understanding of themselves [1]. The needand use of data visualization on mobile devices(e.g., tablets, smartphones, smartwatches) is in-creasingly prevalent in practice, enabling peopleto access data without relying on a desktop orlaptop [2]. Advancements in both software andhardware technologies give us opportunities toprovide people with disabilities with greater ac-cess to data visualization. Multimodal interactionoffers many potential benefits for data visualiza-tion [3], where conventional input modalities suchas keyboard and mouse are not available (e.g.,on mobile devices) or when users have limitedcapabilities (e.g., people with low visualizationliteracy, blind people).

We do not claim that these four topics arecomprehensive or more important than others inreaching broader audiences. However, we positthat they each can enrich the value of visualiza-tion for broader audiences, and more importantlythey together can generate synergy to enhance thecapabilities of the general public to better accessand utilize data. They also suggest a wealth ofimportant and interesting research questions forthe visualization community.

We first describe the characteristics of thesebroader audiences. We then discuss why eachof our four topics deserves more efforts fromthe visualization community, along with someongoing (or stagnant) efforts and recent projects.

CHARACTERIZING BROADERAUDIENCES

Going beyond data-savvy people such as sci-entists and researchers that traditional visualiza-tion research has mainly targeted, our potentialaudiences include people who have the followingcharacteristics. First, they have a wide rangeof data and visualization literacy. While somedata enthusiasts are familiar with more sophis-ticated charts (e.g., scatterplots, node-link dia-grams, treemaps), others may only have a basicknowledge of simple charts (e.g., bar charts, linecharts, pie charts). Lack of prior experience withaccessible charts is a particular issue for peoplewho are blind.

Our broader audiences are likely to have dif-ferent usage contexts and goals. They may notcare about gaining insights from a detailed data

exploration but use data visualization for morecasual purposes [4] such as for fun and curiosity,as well as for simple data access. On the otherhand, some people who are not formally trainedin data analysis or statistics may still need to usedata visualizations to fulfill their tasks.

As these broader audiences expand the usagecontexts, data access environments will vary aswell, going beyond the stationary desktop envi-ronment. People will use data visualization on thego or while lying down on a sofa, both likelyusing mobile devices.

These novel characteristics and contexts poseunique challenges and immense opportunities forvisualization researchers, which we discuss in thefollowing sections.

PERSONAL DATA VISUALIZATIONPeople nowadays have many opportunities to

collect and interact with data due to easy accessto low-cost sensors, mobile devices, and applica-tions. From the data collected with these devicesand applications, people can learn and understandvarious aspects of their lives—such as health,social connectedness, and productivity. Enhancedawareness of one’s own behaviors can be criticalin setting and achieving diverse personal goals.

In the visualization and human-computer in-teraction research communities, researchers havebeen working to help people better collect dataand identify personal insights. Called PersonalInformatics, this growing field of research dealswith supporting ways in which we can leveragepersonal data, by understanding people’s trackingneeds and usage, and by designing, developing,and evaluating new technologies. This field inter-sects with the visualization community in the goalof making it easy for people to better understandtheir data and communicate personal insights. In2014, Huang et al. published a survey paper onpersonal visualization and personal visual ana-lytics, mapping the personal informatics workthrough the lens of visualization research [1].

At VIS 2015, researchers interested inthis topic held a workshop, “Personal Visu-alization: Exploring Data in Everyday Life,”(www.vis4me.com/personalvis15) with a longterm goal of empowering people to makeimprovements to their lives and communitiesthrough advances in personal visualization. Fol-

2

Page 3: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

Figure 1. Examples of data physicalization with personal data [5] (Courtesy of Alice Thudt). (Left) A beadornament daily to decorate branches with 4-hour interval represented by one bead with size showing her moodand colour showing whether she was active, social, or home. (Right) a participant’s six different activities (e.g.,meditation, work) with different coloured beads, each representing one hour, with their size showing enjoyment.

lowing this workshop, however, only a limitednumber of researchers have continued to workat the intersection of visualization and personalinformatics (e.g., data physicalization shown inFigure 1 [5], research methods in personal vi-sualization [6]). We call for further efforts andcontributions to empower the general public whileaddressing meaningful visualization challenges.

To advance and reach the full potential ofpersonal data visualization, we need to exploreways to improve the general public’s visualiza-tion literacy. Although visualizations are heavilyused even in the elementary school curricula, thisdoes not necessarily mean that basic visualizationskills are being taught [7]: Chevalier et al. foundthat teachers perceive visualizations as “intuitive”and therefore do not actively teach them. Visu-alization researchers should continue to designand develop novel visualization techniques for thegeneral public. However, such efforts would bemeaningful only if a certain level of visualizationliteracy is assured.

Going beyond improving basic visualizationliteracy for the general public, there are caseswhere people can benefit from being able toexplore data. For example, dedicated self-trackers(also known as Quantified Selfers) commonlycollect and explore data, and share interestinginsights drawn from their data [8]. However, amajority of people find it challenging to construct

appropriate visualizations, which involves trans-lating questions into data attributes and choos-ing visual mappings [9]. Such skills should bemore easily attainable for people who wantto go a step further in exploring their data—either through teaching existing tools such asTableau (www.tableau.com) and Microsoft PowerBI (www.powerbi.com) or innovating new toolsthat can make the visualization construction andexploration process much easier to learn and use.

To design effective personal visualization, de-signers and researchers should better understandthe personal data context to inform the visualiza-tion design. These contexts are often diverse, on-the-go, situated, and personalized. Creative waysand methods, such as leveraging wearable cam-eras to capture a variety of in-situ contexts [10],are being developed by and introduced to the re-search community. Visualization researchers mayleverage a variety of in-situ study methods bothfor formative and evaluative purposes that wouldsuit personal data contexts. We encourage vi-sualization researchers to proactively collaboratewith the broader HCI communities to infuse anddiscuss design and UX aspects.

DATA VISUALIZATION ON MOBILEDEVICES

Smartphone ownership has surpassed per-sonal computer (i.e., PC and laptop) ownership

March/April 2020 3

Page 4: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

Visualization Viewpoints

(tinyurl.com/qnhbf7b). Smartphones are beingused for a broad range of activities where visual-ization can help to provide easier access to data.This involves data delivered from specific appson smartphones and tablets related to personallyrelevant data such as weather, personal finances,fitness tracking, and health monitoring, as wellas to data related to business- and work-relatedinformation including calendars, measurementstaken, and products sold. Large journalism outletshave been producing data-driven stories with theirmobile readership in mind or even using a mobile-first design concept. In addition, many mobileapps include simple data visualizations with addi-tional goals to attract audiences—fitness trackingapps are a prime example (Figure 2).

The accumulated knowledge in the visualiza-tion community, however, only helps the design-ers of mobile apps and responsive websites to alimited extent. As visualization research in thepast focused on desktop environments, it is oftenunclear if guidelines will hold for the fundamen-tally different mobile environments with changedphysical display sizes, interaction mechanisms,target audiences, and usage contexts. For exam-ple, we need to rethink the traditional WIMP(windows, icons, menus, pointer) interface viadirect manipulation for mobile devices. Mobilevisualizations cannot rely on precise selections orlarge cascading menus for data- and visualization-related operations. In addition, common interac-tion modes such as keyboard shortcuts to selectmultiple data points, right-clicks to choose adesired operation, and mouse-over to show detailson demand become unavailable. Thus, mobile vi-sualization requires new approaches to interactiondesign and potentially needs to be operable witha reduced number of manipulations for makingchanges to data, representation, presentation, andview parameters.

The difficulties of designing visualizations formobile, however, are not only related to the inter-action concept. Mobile devices these days oftenhave very high-resolution displays but the limitedphysical size requires rethinking minimal sizesfor the display of data points and reference struc-tures (e.g., grids, labels, tickmarks). Scalabilityregarding smaller rather than larger display sizesbecomes a problem for mobile visualizations andwe need more perceptual studies to understand

Figure 2. The Fitbit mobile app shows daily sleepscores with a standard bar chart (left) while displayingdifferent sleep stages with a custom chart (right).

the effects of minimizing different types of visu-alizations and their reference structures.

A further design challenge for mobile devicesarises from varying and unconventional usagecontexts compared to desktop computer use. As inthe case of personal data contexts, mobile devicesare used on the go, with moving observers andunder varying lighting conditions. Quick informa-tion access—at a glance—becomes a new usagecontext for which there is little concrete designadvice. Although there has been some research onambient or peripheral displays in the broader HCIcommunity, it would be important to explore ifand how they transfer to the visualization context.

Recently, there have been two efforts in es-tablishing a community around mobile data visu-alization: (1) the “Data Visualization on MobileDevices” workshop at CHI 2018 [2] (mobile-vis.github.io) and (2) the “Mobile Data Visu-alization” Dagstuhl Seminar in July 2019 [11](tinyurl.com/mobilevis-dagstuhl). We call for vi-sualization researchers to attract researchers fromrelated communities to advance this importantand exciting topic.

INCLUSIVE DATA VISUALIZATIONApproximately 36 million people in the world

are blind and an additional 217 million havemoderate to severe vision impairment. This largecohort is currently excluded from the benefits of

4

Page 5: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

(a) (b) (c)Figure 3. Possible ways of presenting data to blind users. (a) Traditionally spatial data is presented using tactilegraphics. (b) A recent approach is to present data on a touchscreen with audio and sometimes haptic feedback.(c) A more futuristic possibility is to use dynamic physical displays [12] (Courtesy of Jason Alexander).

data visualization. While blindness or low-vision(BLV) is the most obvious kind of disabilityaffecting access to data visualization, people withintellectual and developmental disabilities (IDDs)may also find it difficult to use existing visual-ization tools because of impaired cognitive pro-cessing [13], and people with physical disabilitiesmay find it difficult to interact with them.

As the importance of data visualization in-creases and it is more widely used for data-driven reasoning and communication, this raisesa serious equity issue, one that the data visu-alization community has almost totally ignored(one exception is color blindness). Furthermore,improving accessibility is not just of benefit topeople with disabilities, but to all people. Flexi-ble voice-activated solutions, for example, couldallow people to use applications in a hands-freemode while cooking, crafting, or doing sports.

One consideration when designing more in-clusive data visualization applications is thechoice of interaction modalities. Clearly the moreflexibility in input modality, the more accessiblethe application will be. As an indication of thechallenge think about someone who does nothave use of their limbs. How can they controla data visualization application? Is it possible toprovide a natural language interface with eye-gaze controlled object and region selection?

An even more fundamental rethink of datavisualization is required if we wish to provideaccess to people who are blind. Data must nowbe communicated through sound or touch, ratherthan vision (see Figure 3). The most commonway to present graphical information to someone

who is blind is by providing an audio or brailledescription. However, while this can provide ahigh-level overview it necessarily loses detailedinformation and does not allow the reader todevelop their own independent interpretation.

Accessibility guidelines recommend the useof raised line drawings, called tactile graphics,for graphics in which spatial relationships areimportant, such as maps, charts, diagrams, andgraphs. Studies have shown that tactile graphicsprovide touch readers with many of the benefitsthat visualizations provide sighted readers. Forexample, a recent study compared a tactile node-link diagram with a braille list representation ofa network [14]. The blind participants found thenode-link diagram intuitive and were faster andmade less errors with this representation whencounting the number of clusters or finding a pathbetween two nodes.

Unfortunately, traditional tactile graphics arenot well-suited to interactive exploration of dataas each tactile graphic costs around US $1 toprint. Data exploration requires a refreshabledisplay. While refreshable braille displays usingpeizo-electric actuated pins have been availablefor many years, currently larger displays suitablefor displaying graphics using this technology areprohibitively expensive. Fortunately, low-cost re-freshable braille displays using other approachesare about to become available, potentially allow-ing their use for interactive data visualization.

Data physicalization is a promising way ofcreating more accessible visualizations. The visu-alization community has started to explore phys-ical data representations as a way of creating

March/April 2020 5

Page 6: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

Visualization Viewpoints

more engaging presentations (e.g., Figure 1), andthe accessibility community is gaining consider-able interest in the use of 3D printing to createaccessible models for use in education and ori-entation and mobility training for BLV people.Commodity 3D printing now allows 3D models tobe produced at the same cost as tactile graphics.However, standard 3D models have the samedisadvantage as tactile graphics: they are static.Recently there has been research into dynamicdata physicalization (e.g., Figure 3c [12]) butmore research is required for these devices tobecome practical.

Inclusive data visualization raises a large num-ber of research issues. These range from thedesign of new display technologies such as low-cost dynamic physical displays and new interfacesallowing hands-free operation, to the developmentof evidence based guidelines for the design ofvisualizations that are suitable for low vision. Weparticularly note the lack of research investigatingintellectual and developmental disabilities (IDDs)and data visualization [13]. Research is requiredto investigate the impact of IDDs on understand-ing and how to design visualizations that are moreaccessible to this community. In general, makingdata visualization more accessible to people withvarious disabilities will require visualization re-searchers to partner with the accessibility commu-nity to ensure that the solutions developed reallydo meet the needs of the intended users.

MULTIMODAL INTERACTION FORDATA VISUALIZATION

The majority of visualization applications de-veloped to date, particularly those designed foranalysts and experts, have provided a singlemethod of interaction, namely direct manipulationvia a mouse and keyboard. A multimodal inter-face, conversely, utilizes multiple modes of inputsuch as touch, pen, speech, natural language, andgesture to allow a person to interact with andcontrol the system.

To investigate the potential benefits of lever-aging multiple input modalities, several visualiza-tion systems have combined at least two modes ofinput. For example, SketchInsight allows peopleto visually explore their data by interactivelycreating and manipulating charts with pen andtouch interaction [16]. More recently, Orko (Fig-

ure 4) combines speech-based natural languagewith touch-based direct manipulation to facilitatenetwork-based data exploration [15].

These alternative methods of input, sometimescharacterized as Post-WIMP [17], can help visu-alization reach the new audiences we have identi-fied in the earlier section. Multimodal interactionhelps us tackle the fundamental challenges ininteraction design when a mouse and keyboardare not available, and when the display device issmaller and/or movable. This includes the casewhere a visualization is presented on a mobiledevice, as discussed earlier. Speech commandscan replace operations typically entered througha keyboard or manipulated through a mouse,and they allow the removal of control panelsthat take up valuable display real estate. Forexample, speech interaction can enable easy dataentry (e.g., date, time, date ranges), which isoften tedious to do on mobile devices [18]. Suchinterfaces could allow people to easily exploretime-oriented data (e.g., self-tracking data) onmobile devices.

We envision other scenarios where multi-modal interaction could help bring visualizationto novices. Natural language interaction, espe-cially when combined with direct manipulation,can lower the barrier to access data by makingit easy to express data-oriented questions with-out learning native database query languages orspecific user interfaces [19]. Such interfaces couldallow people to express questions about data morenaturally, as they would in everyday conversation.

Multimodal interaction can also be used tomake data visualization more accessible to peoplewith disabilities. For example, one approach tosupport interactive exploration or consumption ofdata by BLV people is to present visualizationon a touchscreen device using a mix of touchtriggered audio and haptic feedback where theaudio feedback includes both sonification and au-dio description (see Figure 3b). This approach hasbeen explored for accessible graphing calculators(e.g., [20]), but to the best of our knowledge hasnot been employed and evaluated more broadlyfor interactive exploration or consumption of data.Speech input also harbors great potential to assistpeople whose motor disabilities make the useof traditional input devices such as a mouse,keyboard, and pen difficult or impossible.

6

Page 7: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

Figure 4. Orko is a network visualization tool that facilitates multimodal interaction by combining naturallanguage (with speech) and direct manipulation (with touch) [15].

The “Multimodal Interaction for Data Vi-sualization” workshop at AVI 2018 [3] (multi-modalvis.github.io) is the initial attempt to builda community of multimodal visualization re-searchers and establish an agenda for researchon multimodal interactions for and with visual-ization. We call for follow-up efforts from thevisualization research community to advance thistopic, investigating important and interesting re-search questions.

OTHER RESEARCH TOPICSWe acknowledge that other topics beyond

these four can help visualization reach broaderaudiences. Some of them are covered by otherviewpoints articles or mentioned earlier in thisarticle. For example, visualization literacy [7] isan important topic. It is critical underpinningfor personal data visualization and also playsa role in increasing accessibility. Data-drivenstorytelling [21] has the potential to improvecommunication of information to targeted audi-ences. Data physicalization can encourage affec-tive, physical, intellectual, and social engagementfor various target groups [22]. In addition, re-search into visualization for and with AI, couldhelp to understand and target the needs of wideruser groups. We hope to see other interesting top-ics emerge, of course, also around topics outsidethe ones we focused on.

CONCLUSIONThe field of information visualization has

made great strides, empowering experts to dealwith massive data leveraging visualization in theirwork. However, we note that visualization re-search falls short on serving a larger group ofpeople who can also benefit from data visualiza-tion. As they have a wider range of data andvisualization literacy as well as more diverseusage contexts and goals, these broader audiencespose interesting challenges and opportunities forvisualization researchers. We specifically calledfor four research topics—personal data visualiza-tion, data visualization on mobile devices, inclu-sive data visualization, and multimodal interac-tion for data visualization: they would help usreach broader audiences and increase the practicalimpact of data visualization.

REFERENCES1. D. Huang, M. Tory, B. A. Aseniero, L. Bartram, S. Bate-

man, S. Carpendale, A. Tang, and R. Woodbury, “Per-

sonal visualization and personal visual analytics,” IEEE

Transactions on Visualization and Computer Graphics,

vol. 21, no. 3, pp. 420–433, 2014.

2. B. Lee, M. Brehmer, P. Isenberg, E. K. Choe,

R. Langner, and R. Dachselt, “Data visualization on

mobile devices,” in Extended Abstracts of the 2018 CHI

Conference on Human Factors in Computing Systems.

ACM, 2018, p. W07.

March/April 2020 7

Page 8: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

Visualization Viewpoints

3. B. Lee, A. Srinivasan, J. Stasko, M. Tory, and V. Setlur,

“Multimodal interaction for data visualization,” in Pro-

ceedings of the 2018 International Conference on Ad-

vanced Visual Interfaces. ACM, 2018, p. 11.

4. Z. Pousman, J. Stasko, and M. Mateas, “Casual in-

formation visualization: Depictions of data in everyday

life,” IEEE Transactions on Visualization and Computer

Graphics, vol. 13, no. 6, pp. 1145–1152, Nov. 2007.

5. A. Thudt, U. Hinrichs, S. Huron, and S. Carpendale,

“Self-reflection and personal physicalization construc-

tion,” in Proceedings of the 2018 CHI Conference on

Human Factors in Computing Systems. ACM, 2018, p.

154.

6. A. Thudt, B. Lee, E. K. Choe, and S. Carpendale, “Ex-

panding research methods for a realistic understanding

of personal visualization,” IEEE computer graphics and

applications, vol. 37, no. 2, pp. 12–18, 2017.

7. F. Chevalier, N. Henry Riche, B. Alper, C. Plaisant,

J. Boy, and N. Elmqvist, “Observations and reflections

on visualization literacy in elementary school,” IEEE

computer graphics and applications, vol. 38, no. 3, pp.

21–29, 2018.

8. E. K. Choe, B. Lee, and m. c. schraefel, “Characterizing

visualization insights from quantified selfers’ personal

data presentations,” IEEE computer graphics and appli-

cations, vol. 35, no. 4, pp. 28–37, 2015.

9. L. Grammel, M. Tory, and M.-A. Storey, “How informa-

tion visualization novices construct visualizations,” IEEE

Transactions on Visualization and Computer Graphics,

vol. 16, no. 6, pp. 943–952, 2010.

10. S. Pizza, B. Brown, D. McMillan, and A. Lampinen,

“Smartwatch in vivo,” in Proceedings of the 2016 CHI

Conference on Human Factors in Computing Systems.

ACM, 2016, pp. 5456–5469.

11. E. K. Choe, R. Dachselt, P. Isenberg, and B. Lee,

“Mobile data visualization (dagstuhl seminar 19292).”

Schloss Dagstuhl-Leibniz-Zentrum fuer Informatik,

2019.

12. F. Taher, Y. Jansen, J. Woodruff, J. Hardy, K. Hornbæk,

and J. Alexander, “Investigating the use of a dynamic

physical bar chart for data exploration and presenta-

tion,” IEEE Transactions on Visualization and Computer

Graphics, vol. 23, no. 1, pp. 451–460, 2016.

13. K. Wu, S. Tanis, and D. Szafir, “Designing communica-

tive visualization for people with intellectual develop-

mental disabilities,” in Workshop on Visualization for

Communication (VisComm). OSF Preprints, 2019.

14. Y. Yang, K. Marriott, M. Butler, C. Goncu, and L. M.

Holloway, “Tactile presentation of network data: Text,

matrix or diagram?” in Proceedings of the 2020 CHI

Conference on Human Factors in Computing Systems.

ACM, 2020, in press.

15. A. Srinivasan and J. Stasko, “Orko: Facilitating mul-

timodal interaction for visual exploration and analysis

of networks,” IEEE Transactions on Visualization and

Computer Graphics, vol. 24, no. 1, pp. 511–521, 2018.

16. B. Lee, G. Smith, N. Henry Riche, A. Karlson, and

S. Carpendale, “Sketchinsight: Natural data exploration

on interactive whiteboards leveraging pen and touch

interaction,” in 2015 IEEE Pacific Visualization Sympo-

sium (PacificVis). IEEE, 2015, pp. 199–206.

17. B. Lee, P. Isenberg, N. Henry Riche, and S. Carpen-

dale, “Beyond mouse and keyboard: Expanding de-

sign considerations for information visualization inter-

actions,” IEEE Transactions on Visualization and Com-

puter Graphics, vol. 18, no. 12, pp. 2689–2698, 2012.

18. E. K. Choe, B. Lee, and S.-W. Hwang, “Personal data

exploration with speech on mobile devices,” in AVI 2018

Workshop on Multimodal Interaction for Data Visualiza-

tion, 2018.

19. V. Setlur, S. E. Battersby, M. Tory, R. Gossweiler, and

A. X. Chang, “Eviza: A natural language interface for

visual analysis,” in Proceedings of ACM UIST ’16, 2016,

pp. 365–377.

20. C. Goncu and K. Marriott, “Gracalc: An accessible

graphing calculator,” in Proceedings of the 17th Inter-

national ACM SIGACCESS Conference on Computers

& Accessibility. ACM, 2015, pp. 311–312.

21. B. Lee, N. H. Riche, P. Isenberg, and S. Carpendale,

“More than telling a story: Transforming data into vi-

sually shared stories,” IEEE computer graphics and

applications, vol. 35, no. 5, pp. 84–90, 2015.

22. Y. Wang, A. Segal, R. Klatzky, D. F. Keefe, P. Isenberg,

J. Hurtienne, E. Hornecker, T. Dwyer, and S. Barrass,

“An emotional response to the value of visualization,”

IEEE computer graphics and applications, vol. 39, no. 5,

pp. 8–17, 2019.

Bongshin Lee is a Sr. Principal Researcherat Microsoft Research. Contact her at [email protected].

Eun Kyoung Choe is an Assistant Professorin the College of Information Studies at the Uni-versity of Maryland, College Park. Contact her [email protected].

Petra Isenberg is a Research Scientist in visualiza-tion with Inria. Contact her at [email protected].

8

Page 9: Reaching Broader Audiences with Data Visualization · visualization, data visualization on mobile devices, inclusive data visualization, and multimodal interaction for data visualization—that,

Kim Marriott is a Professor at Monash University.Contact him at [email protected].

John Stasko is a Regents Professor in theSchool of Interactive Computing and the Directorof the Information Interfaces Research Group atthe Georgia Institute of Technology. Contact him [email protected].

March/April 2020 9