Post on 02-Apr-2022
HAL Id: hal-01516470https://hal.inria.fr/hal-01516470
Submitted on 1 May 2017
HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.
DCPAIRS: A Pairs Plot Based Decision Support SystemEvanthia Dimara, Paola Valdivia, Christoph Kinkeldey
To cite this version:Evanthia Dimara, Paola Valdivia, Christoph Kinkeldey. DCPAIRS: A Pairs Plot Based DecisionSupport System. EuroVis - 19th EG/VGTC Conference on Visualization, Jun 2017, Barcelona, Spain.�hal-01516470�
Eurographics Conference on Visualization (EuroVis), Posters Track (2017)A. Puig Puig and T. Isenberg (Guest Editors)
DCPAIRS: A Pairs Plot Based Decision Support System
Evanthia Dimara1, Paola Valdivia2 and Christoph Kinkeldey1
1 Inria, France2 University of São Paulo, Brazil
5%
37.537.537.537.537.5
37.537.537.537.537.5
Show me remove
University of Copenhagen
academic reputation: 89
employer reputation: 78.9
student / faculty: 99.7
international faculty(%): 67.5
international students(%): 37.5
citations per faculty: 57.4
art sciences: 72.1
engineering: 68.4
life sciences: 80.6
natural sciences: 77.6
social sciences: 74.9
overall: 80.5
attribute no 1: 100
attribute no 2: 100
attribute no 3: 100
attribute no 4: 100
attribute no 5: 100
attribute no 6: 100
attribute no 7: 100
attribute no 8: 100
attribute no 9: 100
attribute no 10: 100
attribute no 11: 100
high
academicacademicacademicacademicacademicreputationreputationreputationreputationreputation
high
employeremployeremployeremployeremployerreputationreputationreputationreputationreputation
high
student /student /student /student /student /facultyfacultyfacultyfacultyfaculty
high
internationalinternationalinternationalinternationalinternationalfaculty(%)faculty(%)faculty(%)faculty(%)faculty(%)
high
internationalinternationalinternationalinternationalinternationalstudents(%)students(%)students(%)students(%)students(%)
high
citationscitationscitationscitationscitationsper facultyper facultyper facultyper facultyper faculty
Color the dots Color the dots Color the dots Color the dots Color the dots Color the dots
No colorNo colorNo colorNo colorNo colorNo color
high
artartartartartsciencessciencessciencessciencessciences
high
engineeringengineeringengineeringengineeringengineering
high
lifelifelifelifelifesciencessciencessciencessciencessciences
high
naturalnaturalnaturalnaturalnaturalsciencessciencessciencessciencessciences
high
socialsocialsocialsocialsocialsciencessciencessciencessciencessciences
high
overalloveralloveralloveralloverall
low
attribute noattribute noattribute noattribute noattribute no11111
low
attribute noattribute noattribute noattribute noattribute no22222
low
attribute noattribute noattribute noattribute noattribute no33333
low
attribute noattribute noattribute noattribute noattribute no44444
low
attribute noattribute noattribute noattribute noattribute no55555
low
attribute noattribute noattribute noattribute noattribute no66666
low
attribute noattribute noattribute noattribute noattribute no77777
low
attribute noattribute noattribute noattribute noattribute no88888
low
attribute noattribute noattribute noattribute noattribute no99999
low
attribute noattribute noattribute noattribute noattribute no1010101010
low
attribute noattribute noattribute noattribute noattribute no1111111111
low
attribute noattribute noattribute noattribute noattribute no1212121212
low
attribute noattribute noattribute noattribute noattribute no1313131313
low
attribute noattribute noattribute noattribute noattribute no1414141414
low
attribute noattribute noattribute noattribute noattribute no1515151515
low
attribute noattribute noattribute noattribute noattribute no1616161616
low
attribute noattribute noattribute noattribute noattribute no1717171717
low
attribute noattribute noattribute noattribute noattribute no1818181818
low
attribute noattribute noattribute noattribute noattribute no1919191919
low
attribute noattribute noattribute noattribute noattribute no2020202020
low
attribute noattribute noattribute noattribute noattribute no2121212121
low
attribute noattribute noattribute noattribute noattribute no2222222222
low
attribute noattribute noattribute noattribute noattribute no2323232323
low
attribute noattribute noattribute noattribute noattribute no2424242424
low
attribute noattribute noattribute noattribute noattribute no2525252525
low
attribute noattribute noattribute noattribute noattribute no2626262626
low
attribute noattribute noattribute noattribute noattribute no2727272727
low
attribute noattribute noattribute noattribute noattribute no2828282828
low
attribute noattribute noattribute noattribute noattribute no2929292929
low
attribute noattribute noattribute noattribute noattribute no3030303030
low
attribute noattribute noattribute noattribute noattribute no3131313131
low
attribute noattribute noattribute noattribute noattribute no3232323232
low
attribute noattribute noattribute noattribute noattribute no3333333333
low
attribute noattribute noattribute noattribute noattribute no3434343434
low
attribute noattribute noattribute noattribute noattribute no3535353535
low
attribute noattribute noattribute noattribute noattribute no3636363636
low
attribute noattribute noattribute noattribute noattribute no3737373737
low
attribute noattribute noattribute noattribute noattribute no3838383838
low
attribute noattribute noattribute noattribute noattribute no3939393939
low
attribute noattribute noattribute noattribute noattribute no4040404040
low
attribute noattribute noattribute noattribute noattribute no4141414141
low
attribute noattribute noattribute noattribute noattribute no4242424242
low
attribute noattribute noattribute noattribute noattribute no4343434343
low
attribute noattribute noattribute noattribute noattribute no4444444444
low
attribute noattribute noattribute noattribute noattribute no4545454545
low
attribute noattribute noattribute noattribute noattribute no4646464646
low
attribute noattribute noattribute noattribute noattribute no4747474747
low
attribute noattribute noattribute noattribute noattribute no4848484848
low
attribute noattribute noattribute noattribute noattribute no4949494949
low
attribute noattribute noattribute noattribute noattribute no5050505050
low
attribute noattribute noattribute noattribute noattribute no5151515151
low
attribute noattribute noattribute noattribute noattribute no5252525252
low
attribute noattribute noattribute noattribute noattribute no5353535353
low
attribute noattribute noattribute noattribute noattribute no5454545454
low
attribute noattribute noattribute noattribute noattribute no5555555555
low
attribute noattribute noattribute noattribute noattribute no5656565656
low
attribute noattribute noattribute noattribute noattribute no5757575757
low
attribute noattribute noattribute noattribute noattribute no5858585858
low
attribute noattribute noattribute noattribute noattribute no5959595959
low
attribute noattribute noattribute noattribute noattribute no6060606060
low
attribute noattribute noattribute noattribute noattribute no6161616161
low
attribute noattribute noattribute noattribute noattribute no6262626262
low
attribute noattribute noattribute noattribute noattribute no6363636363
low
attribute noattribute noattribute noattribute noattribute no6464646464
low
attribute noattribute noattribute noattribute noattribute no6565656565
low
attribute noattribute noattribute noattribute noattribute no6666666666
low
attribute noattribute noattribute noattribute noattribute no6767676767
low
attribute noattribute noattribute noattribute noattribute no6868686868
low
attribute noattribute noattribute noattribute noattribute no6969696969
low
attribute noattribute noattribute noattribute noattribute no7070707070
low
attribute noattribute noattribute noattribute noattribute no7171717171
low
attribute noattribute noattribute noattribute noattribute no7272727272
low
attribute noattribute noattribute noattribute noattribute no7373737373
low
attribute noattribute noattribute noattribute noattribute no7474747474
low
attribute noattribute noattribute noattribute noattribute no7575757575
low
attribute noattribute noattribute noattribute noattribute no7676767676
low
attribute noattribute noattribute noattribute noattribute no7777777777
low
attribute noattribute noattribute noattribute noattribute no7878787878
low
attribute noattribute noattribute noattribute noattribute no7979797979
low
attribute noattribute noattribute noattribute noattribute no8080808080
low
attribute noattribute noattribute noattribute noattribute no8181818181
low
attribute noattribute noattribute noattribute noattribute no8282828282
low
attribute noattribute noattribute noattribute noattribute no8383838383
low
attribute noattribute noattribute noattribute noattribute no8484848484
low
attribute noattribute noattribute noattribute noattribute no8585858585
65 70 75 80 85 90 95 100
30 40 50 60 70 80 90 100
10 20 30 40 50 60 70 80 90 100
30 40 50 60 70 80 90 100
40 50 60 70 80 90 100
86
88
90
92
94
96
98
100
65
70
75
80
85
90
95
100
30
40
50
60
70
80
90
100
10
20
30
40
50
60
70
80
90
100
30
40
50
60
70
80
90
100
candidates
no chance
hard to get
too far
Edit label
a)
b)
c)d)
e)
f)
65 70 75 80 85 90 95 100
30 40 50 60 70 80 90 100
10 20 30 40 50 60 70 80 90 100
30 40 50 60 70 80 90 100
40 50 60 70 80 90 100
86
88
90
92
94
96
98
100
65
70
75
80
85
90
95
100
30
40
50
60
70
80
90
100
102030405060708090100
30
40
50
60
70
80
90
100
7777777777
9696969696 high
academicacademicacademicacademicacademicreputationreputationreputationreputationreputation
high
employeremployeremployeremployeremployerreputationreputationreputationreputationreputation
high
student /student /student /student /student /facultyfacultyfacultyfacultyfaculty
high
internationalinternationalinternationalinternationalinternationalfaculty(%)faculty(%)faculty(%)faculty(%)faculty(%)
high
internationalinternationalinternationalinternationalinternationalstudents(%)students(%)students(%)students(%)students(%)
high
citationscitationscitationscitationscitationsper facultyper facultyper facultyper facultyper faculty
high
artartartartartsciencessciencessciencessciencessciences
high
engineeringengineeringengineeringengineeringengineering
high
lifelifelifelifelifesciencessciencessciencessciencessciences
high
naturalnaturalnaturalnaturalnaturalsciencessciencessciencessciencessciences
high
socialsocialsocialsocialsocialsciencessciencessciencessciencessciences
high
overalloveralloveralloveralloverall
low
attribute noattribute noattribute noattribute noattribute no11111
low
attribute noattribute noattribute noattribute noattribute no22222
low
attribute noattribute noattribute noattribute noattribute no33333
low
attribute noattribute noattribute noattribute noattribute no44444
low
attribute noattribute noattribute noattribute noattribute no55555
low
attribute noattribute noattribute noattribute noattribute no66666
low
attribute noattribute noattribute noattribute noattribute no77777
low
attribute noattribute noattribute noattribute noattribute no88888
low
attribute noattribute noattribute noattribute noattribute no99999
low
attribute noattribute noattribute noattribute noattribute no1010101010
low
attribute noattribute noattribute noattribute noattribute no1111111111
low
attribute noattribute noattribute noattribute noattribute no1212121212
low
attribute noattribute noattribute noattribute noattribute no1313131313
low
attribute noattribute noattribute noattribute noattribute no1414141414
low
attribute noattribute noattribute noattribute noattribute no1515151515
low
attribute noattribute noattribute noattribute noattribute no1616161616
low
attribute noattribute noattribute noattribute noattribute no1717171717
low
attribute noattribute noattribute noattribute noattribute no1818181818
low
attribute noattribute noattribute noattribute noattribute no1919191919
low
attribute noattribute noattribute noattribute noattribute no2020202020
low
attribute noattribute noattribute noattribute noattribute no2121212121
low
attribute noattribute noattribute noattribute noattribute no2222222222
low
attribute noattribute noattribute noattribute noattribute no2323232323
low
attribute noattribute noattribute noattribute noattribute no2424242424
low
attribute noattribute noattribute noattribute noattribute no2525252525
low
attribute noattribute noattribute noattribute noattribute no2626262626
low
attribute noattribute noattribute noattribute noattribute no2727272727
low
attribute noattribute noattribute noattribute noattribute no2828282828
low
attribute noattribute noattribute noattribute noattribute no2929292929
low
attribute noattribute noattribute noattribute noattribute no3030303030
low
attribute noattribute noattribute noattribute noattribute no3131313131
low
attribute noattribute noattribute noattribute noattribute no3232323232
low
attribute noattribute noattribute noattribute noattribute no3333333333
low
attribute noattribute noattribute noattribute noattribute no3434343434
low
attribute noattribute noattribute noattribute noattribute no3535353535
low
attribute noattribute noattribute noattribute noattribute no3636363636
low
attribute noattribute noattribute noattribute noattribute no3737373737
low
attribute noattribute noattribute noattribute noattribute no3838383838
low
attribute noattribute noattribute noattribute noattribute no3939393939
low
attribute noattribute noattribute noattribute noattribute no4040404040
low
attribute noattribute noattribute noattribute noattribute no4141414141
low
attribute noattribute noattribute noattribute noattribute no4242424242
low
attribute noattribute noattribute noattribute noattribute no4343434343
low
attribute noattribute noattribute noattribute noattribute no4444444444
low
attribute noattribute noattribute noattribute noattribute no4545454545
low
attribute noattribute noattribute noattribute noattribute no4646464646
low
attribute noattribute noattribute noattribute noattribute no4747474747
low
attribute noattribute noattribute noattribute noattribute no4848484848
low
attribute noattribute noattribute noattribute noattribute no4949494949
low
attribute noattribute noattribute noattribute noattribute no5050505050
low
attribute noattribute noattribute noattribute noattribute no5151515151
low
attribute noattribute noattribute noattribute noattribute no5252525252
low
attribute noattribute noattribute noattribute noattribute no5353535353
low
attribute noattribute noattribute noattribute noattribute no5454545454
low
attribute noattribute noattribute noattribute noattribute no5555555555
low
attribute noattribute noattribute noattribute noattribute no5656565656
low
attribute noattribute noattribute noattribute noattribute no5757575757
low
attribute noattribute noattribute noattribute noattribute no5858585858
low
attribute noattribute noattribute noattribute noattribute no5959595959
low
attribute noattribute noattribute noattribute noattribute no6060606060
low
attribute noattribute noattribute noattribute noattribute no6161616161
low
attribute noattribute noattribute noattribute noattribute no6262626262
low
attribute noattribute noattribute noattribute noattribute no6363636363
low
attribute noattribute noattribute noattribute noattribute no6464646464
low
attribute noattribute noattribute noattribute noattribute no6565656565
low
attribute noattribute noattribute noattribute noattribute no6666666666
low
attribute noattribute noattribute noattribute noattribute no6767676767
low
attribute noattribute noattribute noattribute noattribute no6868686868
low
attribute noattribute noattribute noattribute noattribute no6969696969
low
attribute noattribute noattribute noattribute noattribute no7070707070
low
attribute noattribute noattribute noattribute noattribute no7171717171
low
attribute noattribute noattribute noattribute noattribute no7272727272
low
attribute noattribute noattribute noattribute noattribute no7373737373
low
attribute noattribute noattribute noattribute noattribute no7474747474
low
attribute noattribute noattribute noattribute noattribute no7575757575
low
attribute noattribute noattribute noattribute noattribute no7676767676
low
attribute noattribute noattribute noattribute noattribute no7777777777
low
attribute noattribute noattribute noattribute noattribute no7878787878
low
attribute noattribute noattribute noattribute noattribute no7979797979
low
attribute noattribute noattribute noattribute noattribute no8080808080
low
attribute noattribute noattribute noattribute noattribute no8181818181
low
attribute noattribute noattribute noattribute noattribute no8282828282
low
attribute noattribute noattribute noattribute noattribute no8383838383
low
attribute noattribute noattribute noattribute noattribute no8484848484
low
attribute noattribute noattribute noattribute noattribute no8585858585
h)
g)
Figure 1: DCPAIRS overview: Initial view before the user assigns importance weights (left). Use case scenario described in section 2 (right)
Abstract
Visualizations designed to support multi-attribute decisions often use colors to encode the identity of the attributes. This ap-proach facilitates mapping of attributes across multiple coordinated views but it has certain limitations: colors often commu-nicate semantics (e.g., red stands for “danger”) deemed to influence the user’s preference, and qualitative color palettes areof limited scalability. We are currently developing a tool with an alternative approach, DCPAIRS: a pairs plot based decisionmaking support tool that employs a compact overview of the decision space and uses visual encodings that communicate un-certainty and suboptimal preference elicitation. Instead of encoding the identity of attributes we use colors for user-authoredannotations to support the decision making process. A use case scenario of a prospective undergraduate student choosing auniversity from the “QS world university ranking” dataset illustrates the functionality of the tool.
Categories and Subject Descriptors (according to ACM CCS): Visualization System and Toolkit Design, Scalability Issues, Mul-tidimensional Data
1. Introduction and Background
Decision making can be a complex process involving a high num-ber of attributes to take into account. An example is a real estatepurchase that may be based on price, size, the number of rooms,location (distance to work, schools, public transport, neighborhoodquality), etc. Since there is rarely a single optimal alternative, peo-ple often need to consider trade-offs between different attributes(e.g., price vs. size) for which personal preferences play an impor-
tant role. Visualization systems for decision support assist users infinding the best alternative by letting them compare the alternativesaccording to their preferences [GLG∗13, PSTW∗17, CL04]. Suchpreferences are expressed by assigning a personal importance scoreto each one of the attributes (often called attribute weights).
A critical point in the design of decision support tools is that,apart from representing the choice alternatives (data points), de-sign emphasis is also given to the attributes’ representation. Most
c© 2017 The Author(s)Eurographics Proceedings c© 2017 The Eurographics Association.
Evanthia Dimara, Paola Valdivia & Christoph Kinkeldey / DCPAIRS
decision support tools choose to encode the attribute identity witha distinct color to be able to identify and explore them across mul-tiple views [GLG∗13, PSTW∗17, CL04, AWL∗15].
Color is a straightforward way of a uniform attribute encod-ing, but the number of qualitative color categories is very lim-ited [War12, Mun14] and so is the number of attributes that canbe handled by the decision support systems using this approach. Atthe same time color always implies semantics (e.g. red stands for“danger” [War12]) which can bias the perceived importance of anattribute and affect the quality of decisions [Ber16].
If color is not used to represent attribute identities, its semanticscan instead be utilized to support user-authored annotations. Anno-tations in information visualization are considered as new attributesor missing information that users add to the data [Mun14, BM13].However, even though visualization tools have explored the bene-fits of annotations in visual analysis [ZGB∗17, EB12, CBY10] weare not aware of their use in decision support visualization tools.
We are currently developing DCPAIRS, a visualization tool thatassists decisions involving a high number (up to hundreds) of at-tributes and lets users to annotate alternatives during the decisionprocess.
2. DCPAIRS tool
DCPAIRS is a pairs plot based decision support tool supportinguser-authored annotations. It shows decision alternatives in a gen-eralized pairs plot [EGS∗13] in which quantitative attributes areshown in scatterplots, [Cle93] (Figure 1c) and categorical attributescan be displayed as mosaic plots [HK84] or side-by-side boxplots[Tuk77]. All attributes are encoded in shaded boxes; six in the di-agonal for a detailed view (Figure 1a) and the rest in a pixel mapoverview (Figure 1b). The attributes’ importance weights are en-coded with a continuous gray scale to express that preferences areoften relative tendencies rather than precise values [PSTW∗17]. Inthe following we describe the interactive features of DCPAIRS in ause case scenario illustrating how the use of a pairs plot for attributepairs can provide critical insights, supported by user-authored an-notations (Figure 1e) using a qualitative color scheme [HB03].
We downloaded the “University Rankings” from [lin], contain-ing the “QS world rankings 2013” of 906 institutions. The rankingis based on the weighted sum of the attributes: “academic reputa-tion” (40%), “faculty student ratio” (20%), “citations” (20%), “em-ployer reputation” (10%), “international faculty” (5%), and “inter-national students” (5%) (Figure 1a). The dataset contains additionalattributes not considered in the ranking: performance in“arts”, “hu-manities”, “engineering”, and “life and natural sciences” (Fig-ure 1b). The extra boxes in Figure 1b demonstrate the tool’s scala-bility and are not part of the real dataset.
Use Case: Vangelis, a prospective undergraduate student fromAthens, is searching for universities to apply for. He loads thedataset into DCPAIRS and takes a look at the gray values and sliderposition (Figure 1h) of the attributes to understand which attributesare considered in the default ranking score. The attribute pixel map(Figure 1b) gives a quick overview of their overall number and rel-ative importance. He sees that most emphasis is laid on academicreputation, about half on citations and on the faculty-student ratio,
less on employer reputation, and a lot less on international facultyand students. Even though the subject areas are deemed unimpor-tant (signified by the white color), Vangelis is mainly interested inengineering and, to a lesser extent, in art subjects, so he changesthe respective sliders to the maximum and to a low value.
Having a limited budget for application fees, Vangelis can onlyselect three universities to apply for and has the strategy to choose abit less competitive yet suitable institutions to increase his chancesof acceptance. He moves the threshold slider (Figure 1d) to filter allbut the top 1% institutions assuming that these are the most com-petitive ones and tags them with red and a newly defined label “nochance” (Figure 1e, c) . He moves the slider to 2% and labels theyet untagged ones with orange and “hard to get”. Vangelis finallysets a 5% threshold to see institutions he considers as more realistic.Since most attributes do not seem to be much correlated (Figure 1c)(apart from “international faculty” and “international students”), heunderstands that he needs to carefully consider each one individu-ally. By hovering over the dots, the inspector (Figure 1f) allowsVangelis to retrieve more detailed information, and he tags the in-teresting institutions in green color (label “candidates”). He alsouses the orange and red dots as baseline to identify candidates sim-ilar to the top institutions.
Another aspect Vangelis cares about is to find an institution ata location that will allow him to travel to his family from timeto time. This information is not in the dataset, but he extracts itfrom the institution names. He tags the ones extremely far from hishometown (e.g., in China) with blue and the label “too far”. Usingthe color tags he finds his three favorite “safe” choices: “Universityof Copenhagen”, “University of Manchester” and “Kings College”,but he decides to also apply to "University of Oxford” (in red) try-ing his luck with at least one of top institutions.
Vangelis is browsing the pixel map overview to identify other at-tributes that could affect his decision. Since he is interested in lowtuition fees, he changes the default direction-check box (Figure 1g)of the “tuition fee” attribute (hypothetical attribute, not part of thereal dataset) from a “high" to a “low" direction and drags it to thediagonal detailed view to see it paired with “academic reputation”.He observes that “fee” is correlated to “reputation” in most cases,but that a few top institutions do not follow this trend. Vangelis in-terprets this as a sign that the latter have a friendlier policy towardsstudent’s budget and reconsiders his choices.
3. Status and Future Work
DCPAIRS is a pairs plot based decision making support tool that iscurrently under development. A working prototype of the tool withthe functionality described above already exists. However, its per-formance has to be improved to handle higher amounts (hundreds)of decision alternatives fluently. An important part of the develop-ment will be a user study to assess the advantage of a pairs plotapproach and the potential benefits of providing user-authored an-notations during the decision making process.
4. Acknowledgments
We thank Pierre Dragicevic, Anastasia Bezerianos and our review-ers for their precious feedback on the paper.
c© 2017 The Author(s)Eurographics Proceedings c© 2017 The Eurographics Association.
Evanthia Dimara, Paola Valdivia & Christoph Kinkeldey / DCPAIRS
References[AWL∗15] ASENIERO B. A., WUN T., LEDO D., RUHE G., TANG A.,
CARPENDALE S.: Stratos: Using visualization to support decisions instrategic software release planning. In Proceedings of the 33rd An-nual ACM Conference on Human Factors in Computing Systems (2015),ACM, pp. 1479–1488. 2
[Ber16] BERA P.: How colors in business dashboards affect users’ deci-sion making. Communications of the ACM 59, 4 (2016), 50–57. 2
[BM13] BREHMER M., MUNZNER T.: A multi-level typology of abstractvisualization tasks. IEEE Transactions on Visualization and ComputerGraphics 19, 12 (Dec 2013), 2376–2385. doi:10.1109/TVCG.2013.124. 2
[CBY10] CHEN Y., BARLOWE S., YANG J.: Click2annotate: Automatedinsight externalization with rich semantics. In 2010 IEEE Symposiumon Visual Analytics Science and Technology (Oct 2010), pp. 155–162.doi:10.1109/VAST.2010.5652885. 2
[CL04] CARENINI G., LOYD J.: Valuecharts: analyzing linear modelsexpressing preferences and evaluations. In Proceedings of the workingconference on Advanced visual interfaces (2004), ACM, pp. 150–157. 1,2
[Cle93] CLEVELAND W. S.: Visualizing data. Hobart Press, 1993. 2
[EB12] ELIAS M., BEZERIANOS A.: Annotating bi visualization dash-boards: Needs & challenges. In Proceedings of the SIGCHI Conferenceon Human Factors in Computing Systems (2012), ACM, pp. 1641–1650.2
[EGS∗13] EMERSON J. W., GREEN W. A., SCHLOERKE B., CROWLEYJ., COOK D., HOFMANN H., WICKHAM H.: The generalized pairs plot.Journal of Computational and Graphical Statistics 22, 1 (2013), 79–91.2
[GLG∗13] GRATZL S., LEX A., GEHLENBORG N., PFISTER H.,STREIT M.: Lineup: Visual analysis of multi-attribute rankings. IEEEtransactions on visualization and computer graphics 19, 12 (2013),2277–2286. 1, 2
[HB03] HARROWER M., BREWER C. A.: Colorbrewer. org: an onlinetool for selecting colour schemes for maps. The Cartographic Journal40, 1 (2003), 27–37. 2
[HK84] HARTIGAN J. A., KLEINER B.: A mosaic of television ratings.The American Statistician 38, 1 (1984), 32–35. 2
[lin] Lineup - legatum prosperity index. https://lineup.caleydoapp.org. Accessed: 2017-02-24. 2
[Mun14] MUNZNER T.: Visualization Analysis and Design. CRC Press,2014. 2
[PSTW∗17] PAJER S., STREIT M., TORSNEY-WEIR T., SPECHTEN-HAUSER F., MULLER T., PIRINGER H.: Weightlifter: Visual weightspace exploration for multi-criteria decision making. IEEE transactionson visualization and computer graphics 23, 1 (2017), 611. 1, 2
[Tuk77] TUKEY J. W.: Exploratory data analysis. 2
[War12] WARE C.: Information visualization: perception for design. El-sevier, 2012. 2
[ZGB∗17] ZHAO J., GLUECK M., BRESLAV S., CHEVALIER F., KHANA.: Annotation graphs: A graph-based visualization for meta-analysis ofdata based on user-authored annotations. IEEE Transactions on Visual-ization and Computer Graphics 23, 1 (2017), 261–270. 2
c© 2017 The Author(s)Eurographics Proceedings c© 2017 The Eurographics Association.