Introduction

COMP7507
Visualization & Visual Analytics
What is Visualization?
• The communication of information using
graphical representation.
• “A picture is worth a thousand words.”

2
Text vs Images
• Text/words • Images/pictures
o Sequential process of o Interpreted in parallel by
reading the human perceptual
system
o Limited by the use of o Independent of local
different languages language

a curved line
with every point
equal distance
from the center
An example from Mike Parkinson’s “Do-It-Yourself Billion Dollar Graphics
3
 The Global Datasphere
• We are generating a whole lot of data every second.
• How big is the “Global Datasphere”?
o Prediction by IDC (International Data Corporation) in 2020: grow from 45
Zetabytes (ZB) in 2019 to 175ZB by 2025. [link] 1 ZB = 1 trillion GB

o If you store all 175ZB of data on single layer Blu-ray discs, then you can have
a stack of discs that go to the moon 23 times.

An Infographic for the IDC

prediction in 2014.

4
Data and Visualization
• Why visualization?
o Data needs interpretation (e.g., to explain process and
concepts)

o Visualization serves as an effective tool to assist in

analysis and communication

5
How to visualize?

• Data in tabular form is hard to discover relationship,

trends, patterns
6
Daily Life Examples of Visualization
• Newspaper / Magazine – in articles accompanying
text to present data

Simple & static

7
 Daily Life Examples of Visualization
• Weather chart

Low level specifics

Keep the user in mind.

What do they want?
High level abstraction
8
Daily Life Examples of Visualization

Occlusion problems when too much information needs to be shown.

Use interaction to resolve. 9
Daily Life Examples of Visualization
• Stock market

Complex presentation

Designed for target

users with proper
training

Every visualization has a

purpose
10
Daily Life Examples of Visualization
• Medical Care (Brain MRI 2D)

Need expertise and training for spatial association

11
Daily Life Examples of Visualization
• Medical care (Brain MRI 3D Visualization)

3D graphics
embedded with the
actual data

Changing viewpoint
and object
transparency helps
show the context
[VUmc Amsterdam, http://www.youtube.com/watch?v=wjvDDH-uJ0s ]

12
Daily Life Examples of Visualization
• Medical care (Ultrasound)
2D 3D/4D

https://www.youtube.com/watch?v=8if_IVf0yHw https://www.youtube.com/watch?v=nD_l0EtkNYU

13
Examples of Visualization at Work
• Aerodynamics of a racing car

Color encodes physical

properties/quantities,
e.g., vorticity

Visual assistance
for checking if the
result is as expected

[https://www.youtube.com/watch?v=uMEYeQMvzLc ]

14
Examples of Visualization at Work
• Geophysicist in hunt for oil and gas

Immersive
system with big
displays for
exploratory
visualization
[ https://www.youtube.com/watch?v=hCBIAJlztag ]

15
An Example at HKU CS
• Project WATERMAN
( www.waterman.hku.hk )

o An Internet and
GIS-based water
quality forecast and
management system
for HK
o Uses intensive visualization techniques for 3D
Environmental Impact Assessment

16
Some Questions

• Why can visualization explain better than text?

• Is visualization always effective?
• Given the same data, are there different ways to
visualize it? Which is the best?

17
History of Visualization
• Milestones in the
history of thematic
cartography, statistical
graphics, and data
visualization
• From 6200 BC to
present

[ http://www.datavis.ca/milestones/ ]

18
The French Invasion of Russia in 1812
• French invaded Russia in June 1812,
size of the Grand Army = 422K,
starting from Polish-Russian border

• The Napoleon army (size = 100K)

reached Moscow in Sept

• The army retreated from Moscow in

Oct, and underwent severely cold
A story to be told
weather and lack of supplies

• The army crossed the Berezina River in

Nov with only 27K soldiers left

• Finally back to Poland with only 10K

men remaining
19
The French Invasion of Russia in 1812
• French invaded Russia in June 1812, • Variables in the story
size of the Grand Army = 422K,
o Size of the army
starting from Polish-Russian border
o Location of the army
• The Napoleon army (size = 100K) (2D)
reached Moscow in Sept o Direction of the
• The army retreated from Moscow in army’s movement
Oct, and underwent severely cold (2D)
weather and lack of supplies o Temperature during
the retreat from
• The army crossed the Berezina River in Moscow
Nov with only 27K soldiers left
• How to present in
• Finally back to Poland with only 10K
one single drawing?
men remaining
20
The French Invasion of Russia in 1812
• The classic graphic by Charles Joseph Minard (1781-1870)

21
The Cholera Epidemic in London, 1854
• Dr. John Snow’s plot of over 500 deaths in central
London from Cholera in Sept 1854.
• The plot helped identifying the concentration of
deaths near the Broad Street water pump.

22
Original map by Snow
23
 Pump locations marked
as crosses

Concentration of death
leads to the hypothesis
that the disease can be
spread by water

Outliers can be
identified easily for
further verification of
the hypothesis

A simplified version by E.W. Gilbert 24

The Cholera Epidemic in London, 1854
• Snow used the dot plot, marking each individual death.
o Dots are cluttered
o Death rates are not shown

• Question: If a region is without dots, does it mean that

there is no deaths or that the area is not populated?
o Solution: To show the death rates in a region
o However, this requires the aggregation of data, e.g., to sum up
the population within a certain area.

25
The Cholera Epidemic in London, 1854
• Aggregation may sometimes mask relevant
details, distort the actual data, and generate
misleading signals.

26
Why Visualization is Important?
• provides an ability to comprehend huge amounts of
data
• allows the perception of emergent properties that
were not anticipated (new insight)
• often reveals problems with the data itself quickly
(anomalies)
• facilitates understanding of both large-scale and
small-scale features of the data
• facilitates hypothesis formation

27
What Should a Good Visualization Achieve?
• Show the data
• Induce the viewer to focus on the substance rather than the
methodology, graphic design, the technology of graphic
production, etc.
• Avoid distorting the data
• Present many numbers in a small space
• Encourage the eye to compare different pieces of data
• Reveal the data at several levels of detail, from a broad overview
to the fine structure
• Serve a reasonably clear purpose: description, exploration,
tabulation, or decoration
28
Some Bad Examples
• Data:
o the percentage of college students in US that were
under 25 from 1972 through 1976

Year Percentage
1972 72.0
1973 70.8
1974 67.2
1975 66.4
1976 67.0

29
Some Bad
Examples
Tufte (1983, p.118) says,
“This may be the worst
graphic ever to find its
way into print.”

Simple guideline:

• Use table to show small data set with

< 20 numbers

• Use graphic for large data set

30
Some Bad Examples

Which state in US has the highest % of homes exceeding the recommended Radon
level? Which one has the lowest? [Things that Make Us Smart, p70-71]

31
Some Bad Examples

32
Some Bad Examples
• A much better presentation

33
Graphics Illusion
• Local vs. Global

34
Graphics Illusion

35
Graphics Illusion

36
M.C. Escher (1898-1972)

37
Escher for Real

http://www.cs.technion.ac.il/~gershon/personal/EscherForReal/

38
Graphic that fails to tell the truth
• A graphic does not distort if the visual
representation of the data is consistent with the
numerical representation.
• What is visual representation?
o Physically measured?
o Perceived visual effect?

• One thing to note: Perception changes with

experience and is context-dependent
39
Graphic Distortion

• % increase from 1978 to 1985 = (27.5-18)/18*100% = 53%

• % increase of change in graph = (5.3-0.6)/0.6 = 783%
• Lie factor = size of effect in graphic / size of effect in data = 783 / 53 = 14.8

40
Graphic Distortion
• Can be done as simple as this:

Fuel Economy Standards for Auto

30

25
Miles per Gallon

20

15

10

0
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

Year

41
Graphic Distortion

Looks like a steep

change

Illusion due to 3D
perspectives and the
non zero-based y-axis

42
Design vs. Data Variations
• A design is expected to be consistent over the
entire graphic. E.g., the intervals depicted by an
axis in a graph should be of uniform scale
• Should show data variation, but not design
variation.

43
Design vs. Data Variations

Problems:

Overloads viewers with

two graphics of different
horizontal scales & vertical
scales in a seemingly
integrated chart.

44
Visual Area vs. Numerical Measure

• Using areas to show

one-dimensional
data variation is
confusing.

45
Context is Essential
• Ask: “Compare to what?”

46
Principles of Graphic Integrity
• The representation of numbers, as measured on
the surface of the graphic itself, should be directly
proportional to the numerical quantities
represented.
• Clear, detailed and thorough labeling should be
used to defeat graphical distortion and ambiguity.
• Graphics must not quote data out of context.

47
Visualization vs. Computer Graphics
• Computer graphics
o Digital synthesis and manipulation of visual contents
(geometry/imaging/rendering/animation)
o Visual realism is one of the primary goals
o Big impact in animation/movies/video games
• Visualization
o Applies computer graphics techniques to generate visual display
of data
o Emphasizes on effective communication of information

48
Visualization vs. Computer Graphics
• An example:

Smoke visualization Smoke rendering

[von Funck et al., VIS 2008] [Fedkiw et al., SIGGRAPH 2001]

49
State-of-the-Art Research
• Visualization
o IEEE VIS (InfoVis / SciVis) (ieeevis.org)

• Visual Analytics
o IEEE VIS (VAST) (ieeevis.org)
o Visual Analytics Community (vacommunity.org)

• Computer Graphics
o SIGGRAPH (siggraph.org)

50
References
• Matthew Ward, Georges Grinstein and Daniel Keim,
"Interactive Data Visualization: Foundations, Techniques, and
Applications", 2010 [Chapter 1]

• Edware R. Tufte, “The Visual Display of Quantitative

Information”, 2001.

• D. Keim, J. Kohlhammer, G. Ellis and F. Mansmann, “Mastering

the Information Age: Solving Problems with Visual Analytics”,
2010.

51