Slides - Interaction

Interaction 

Jing Yang 

Univ. North Carolina at Charlotte 

What is Interaction 

• From Google: Reciprocal action between a 

human and a computer 

• One of the two main components in infovis 

• Representation 

• Interaction 

• Interaction is what distinguishes infovis from 

static ti visual representations ti on paper 

2 

1

Interaction Types 

• Keim’s taxonomy (TVCG ’02) includes 

• Projection 

• Filtering 

• Zooming 

• Distortion 

• Linking and brushing 

3 

Interaction Types 

• Dix and Ellis (AVI ’98) propose 

• Highlighting hti and focus 

• Accessing extra info – drill down and 

hyperlinks 

• Overview and context – zooming and fisheyes 

• Same representation, changing parameters 

• Linking representations ti – temporal fusion 

4 

2

In this class 

• Interactions for browsing a large amount of 

data 

• Labeling, animations, overlap reduction 

techniques, multiple views 

• Interactions for selecting data of interest 

• Brushing, filtering, dynamic query 

• Interactions ti for examining i data of interest 

t 

• Zooming, distortion, roll up/drill down 

5 

In this class 


data 




• Interactions for examining data of interest 

6 

3

Labeling 

• Why labels are needed in visualizations 

• To associate visualization with data 

• To display semantics 

• To explain data, visualization, and 

relationships 

• Labeling Challenges [Fekete and Plaisant CHI99] 

• Readable 

• Non-ambiguously related to its graphical 

object 

• Does not hide any pertinent information. 

7 

Taxonomy of labeling… 

“Excentric Labeling: Dynamic Neighborhood Labeling for Data 

Visualization”, Jean-Daniel Fekete, Catherine Plaisant, CHI 99 

8 

4

Dynamic Labeling (1) 



9 




10 

5




11 

http://www.excelcharts.com/blog/focuscontext-bar-chart-skyscraper/ 

12 

6




13 

Excentric Labeling 

Demo 

http://www.cs.umd.edu/hcil/excentric/#prototypes 



14 

7

Evaluation of Excentric Labeling 

• Comparison of excentric 

with virtual instantaneous 

zoom. 

• a 60% speed advantage 

for the excentric 

• Easily learnable after a 

little practice. 

• No of operations in zoom 

was much more 



15 

Algorithm 

• 1. Extract each label and position for interesting 

graphic objects in the focus region. 

• 2. Compute an initial position. 

• 3. Compute an ordering. 

• 4. Assign the labels to either a right or left set. 

• 5. Stack the left and right labels according to their 

order. 

• 6. Minimize the vertical distance of each set from the 

computed initial position. 

• 7. Add lines to connect the labels to their related 

graphic object. 



16 

8

Particle-Based Labeling 

• Optimized label all 

• Fast 

• In arbitrary shapes 

• Video 

“Particle-Based Labeling: Fast Point-Feature Labeling 

without Obscuring Other Visual Features”, Martin 

Luboschik, Heidrun Schumann, Hilko Cords, InfoVis 

2008 

17 

Use the technique carefully 

18 

9

Dynamic Graph Labeling 

• Target: graphs with extended node and link 

labels, whose lengths range from a short 

phrase to a full sentence to an entire 

paragraph and beyond. 

• Solution: share visualization resources with 

graph and present label in static, interactive, 

and dynamic mode. 

“Dynamic Visualization of Graphs with Extended 

Labels”, Pak C. Wong et al. Infovis 2005 

19 

Label -> Link 

20 

10

Directed graph 

21 

Alpha Blending 

22 

11

Node Labels 

Video 

23 

In this class 


data 





24 

12

Animations 

Why animations 

• To show trends 

• To add a smooth transition that relates two 

displays 

• To add more information into constrained 

displays 

• To reduce clutter 

• To reveal relationship 

25 

Animations for Showing Trends 

• Gapminder Trendalyzer - animated bubble 

chart to show trends over time in three 

dimensions 

• Video: Hans Rosling, No more boring data. 

TED (Technology, Entertainment, Design) 

2006 

• Demo: Gapminder World 

26 

13

Animations for Smooth Transitions 

• A commonly held belief 

• Animation helps users maintain i object 

constancy and thus helps users to relate 

the two states of the system 

• A reported user study [Bederson and Boltman 

Infovis99]: 

• Increased users’ ability to reconstruct t info 

space 

• No penalty on task performance 

• Cost extra in response time vs. Relate two 

27 

states fasters 

Rolling the dice 

• Elmqvist, N.; Dragicevic, P.; Fekete, J.-D., 

Rolling the Dice: Multidimensional Visual 

Exploration using Scatterplot Matrix 

Navigation Visualization, InfoVis 2008 

• Video 

28 

14

Animations for More Information 

• Video: Sophie Engle et al., Cenimation – 

Visualization for constrained displays 

29 

Animations for Reducing Clutter 

• Video: EdgeLens [Wong at. el. Infovis 03], 

http://grouplab.cpsc.ucalgary.ca/papers/videos/ 

ucalgary ca/papers/videos/ 

30 

15

Animation for Revealing Relationship 

• Video: Dust & Magnet: Multivariate 

Information Visualization using a Magnet 

Metaphor, Yi et al. Information Visualization 

(2005) 

31 

In this class 


data 





32 

16

Overlap 

Reduction 

• Extent Scaling 

• Dynamic Masking 

• Zooming and Panning 

• Showing Names 

• Layer Reordering 

• Manual Relocation 

• Automatic Shifting 

Value and Relation Display for Interactive 

Exploration of High Dimensional Datasets, J. 

Yang et al. InfoVis 2004 

SkyServer Dataset 

33 

In this class 


data 





34 

17

Changing Representation 

• May interactively change entire data 

representation 

• Looking for new perspective 

• Limited real estate may force change 

35 

Rearranging View 

• Keep same fundamental representation and 

what data is being shown, but rearrange 

elements 

• Alter positioning 

• Sort 

Demo: XmdvTool 

36 

18

Sorting in Multidimensional 

Visualization 

• Can sort data with respect to a particular 

attribute in Table Lens 

37 

Sorting in Graph Visualization 

• rank nodes by analytical information such as 

centrality, degrees... 

• use ordered list, scatterplots, visually coded 

node-link diagrams to provide overview, filter 

nodes, and find outliers 

• aggregate ranking for cohesive subgroups 

“Balancing Systematic and Flexible Exploration of 

Social Networks”. Adam Perer and Ben 

Shneiderman, InfoVis 06 

38 

19

1-D Ranking 

39 

2-D Ranking 

40 

20

Overview + Detail View 

• Example: Visible Human Explorer - video 

http://www.nlm.nih.gov/research/visible/vhp 

_conf/north/vhedemo.htm 

41 

Jigsaw 

• http://www.cc.gatech.edu/gvu/ii/jigsaw/ 

• Video 

42 

21

In this class 


data 




43 

Brushing 

• Brushing - Selecting or highlighting a case in 

one view generates highlighting the case in 

the other views 

• Often used when there are multiple views of 

the same data – coordinated views 

• Viewer may wish to examine different 

attributes of a data case simultaneously 

• Alternatively, viewer may wish to view data 

case under different perspectives or 

representations 

• But need to keep straight where the data case 

is 

44 

22

N-D Brushing (demo) 

45 

Structure-Based Brushing (demo) 

46 

23

In this class 


data 




47 

Filtering/Limiting 

• Changing the set of data cases/dimensions 

being presented 

• Focusing 

• Narrowing/widening 

• Example: Faceted search 

48 

24

In this class 


data 




49 

Dynamic Query 

• DB Queries 

Select house-address 

From atl-realty-db 

Where price >= 200,000 and 

price = 3 

50 

25

Typical Query Response 

• 124 hits found 

• 1. 748 Oak St. - a beautiful … 

• 2. 623 Pine Ave. - 

• … 

• 0 hits found 

51 

Problems 

• Must learn language 

• Only shows exact matches 

• Don’t know magnitude of results 

• No helpful context is shown 

• Reformulating to a new query can be slow 

• ... 

52 

26

Dynamic Query 

• Specifying a query brings immediate display 

of results 

• Responsive interaction (< .1 sec) with data, 

concurrent presentation of solution 

• “Fly through the data”, promote exploration, 

make it a much more “live” experience 

“Dynamic Queries for Visual Information 

Seeking”, B. Shneiderman, Software, 11(6), 

pages 70-77 

53 

Dynamic Query Constituents 

• Visual representation of world of action 

including both the objects and actions 

• Rapid, incremental and reversible actions 

• Selection by pointing (not typing) 

• Immediate and continuous display of results 



pages 70-77 

54 

27

Idea at heart of Dynamic Query 

• There often simply isn’t one perfect response 

to a query 

• Want to understand a set of tradeoffs and 

choose some “best” compromise 

• You may learn more about your problem as 

you explore 



pages 70-77 

55 

DQ Strengths 

• Work is faster 

• Promote reversing, undo, exploration 

• Very natural interaction 

• Shows the data 

56 

28

Videos 

• 1. Ben’s dynamic query talk 

• 2. Filmfinder 

• 3. Ben’s spotfire talk 

• 4. Dynamap 

57 

In this class 


data 



• Zooming, details-on-demand, distortion, rolling 

up/drilling down 

58 

29

Zooming/Panning 

• Zooming in - the interaction that changes the current 

display from a view of a lower level of detail to a view 

of a higher level of detail. 

• Zooming out - the interaction that changes the 

current display from a view of a higher level of detail 

to a view of a lower level of detail. 

• Panning - the interaction that changes the current 

display from a subregion of a view to an adjacent 

sub-region of the same view. There can be overlaps 

between the two regions. 

59 

Example - MapQuest 

60 

30

Panning and Zooming 

• Panning in high levels of detail can be time 

consuming 

• Solution: zoom out, pan, and zoom in 

• Drawbacks: jitter in the process 

• Improvement: Smooth and Efficient Zooming and 

Panning (van Wijk and Nuij, Infovis 03) 

61 

Panning and Zooming 

• “Speed-Dependent Automatic Zooming for 

Browsing Large Documents” Igarashi & 

Hinckley, Proc. UIST'00, pp. 139-148. 

• Keep constant perceptual scrolling speed 

• Scale X Speed = Constant 

Video! 

62 

31

Zooming and Panning 

• SpaceTree: Supporting Exploration in Large 

Node Link Tree, Design Evolution and 

Empirical Evaluation Grosjean, Plaisant and 

Bederson, InfoVis 2002 

• A zooming environment that dynamically lays 

out branches of a tree to best fit and available 

screen space 

• Video 

63 

SmartSkip 

• Video: “Smart Skip: Consumer Level 

Browsing and Skipping of Digital Video 

Content”, Steven Drucker et al., CHI 2002 

64 

32

In this class 


data 





65 

Details-on-Demand 

• To provide viewers with more information/details 

about data items 

• May just be more information about data items 

• May be moving from aggregation view to individual 

view 

• May not be showing all the data due to scale problem 

• May be showing some abstraction of groups of 

elements 

• Expand set of data to show more details, perhaps 

individual cases 

66 

33

DragMag 

• Colin Ware, Marlon Lewis: The DragMag 

image magnifier. CHI 95 Conference 

Companion 1995: 407-408 

• video 

67 

In this class 


data 

• Labeling, overlap reduction techniques 






68 

34

Distortion 

• Distortion - an operation that increase the 

screen space allocated to some objects in the 

display while decreasing the screen space 

allocated to other objects. 

69 

Magnifier Lens 

Figure from [Robertson & Mackinlay UIST 93] 

70 

35

Magic Lens 

• Video (The Movable Filter as an Interface 

Tool, Bier et al. CHI’95) 

71 

FishEye Lens 

Algorithm: Generalized Fisheye Views, G. Furnas, CHI’86 

72 

36

Fisheye Menus 

• Bederson, B. B. (November 2000) 

Fisheye Menus 

Proceedings of ACM Conference on User 

Interface Software and Technology (UIST 

2000), pp. 217-226, ACM Press. 

• Video 

73 

Table Lens [RC95] 

74 

37

Table Lens Distortion in Scatterplots 

75 

Table Lens [RC95] - video 

76 

38

Datelens 

• Video 

• http://www.cs.umd.edu/hcil/datelens/ 

d /h t l / 

77 

Perspective Wall [MRC91] 

78 

39

Document Lens 

[Robertson & Mackinlay UIST 93] 

79 

Flip Zooming [Holmquist SIGCHI 97] 

80 

40

Flip Zooming 

81 

Hierarchical Image Browser 

[Holmquist and Bjőrk SIGGRAPH 98] 

82 

41

Complex Logarithmic Views for Small 

Details in Large Contexts. [J. Böttger et al. 06] 

• Idea: use the complex logarithm and root 

functions to show very small details even in 

very large contexts (video) 

83 

Reference 

• John stasko’s infovis class slides 

84 

42

Slides - Interaction

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?