VISUAL ENCODING AND PRESENTATION

Bùi Tiến Lên

2024
Contents

1. Marks And Visual Variables

2. Expressiveness and Effectiveness

3. Evaluation

4. Data Visualization Gallery

Marks And Visual Variables
 Marks
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation Concept 1
Data
Visualization Marks are basic geometric elements that serve as the carriers of information.
Gallery
They are distinguished by their dimensionality. There are 0D points, 1D lines, 2D
areas, and 3D bodies.

Marks as Items/Nodes
Points Lines Areas

Marks as Links
Containment Connection

4
 Visual Variables
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation Concept 2
Data
Visualization Visual variables (aesthetics, channels) such as position, shape, or hue, which
Gallery
control appearance of marks.

Position Lightness

Length Saturation

Area Hue

Angle Shape

5
 Visual Variables (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation
Variables Changes
Data
Visualization Position changes in the x, y, location
Gallery
Size changes in length, area
Shape infinite number of shapes
Value (lightness, saturation) changes from light to dark
Hue changes in hue at a given color
Orientation (angle) changes in alignment
Texture variation in ’grain’

6
 Properties of visual variables
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • Selective: A visual variable is said to be selective if a mark changed in this

Data
Visualization
variable alone makes it easier to select that changed mark from all the other
Gallery marks.
• Associative: A visual variable is said to be associative if marks that are like
in other ways can be grouped according to a change in this visual variable.
• Quantitative: A visual variable is said to be quantitative if the relationship
between two marks differing in this visual variable can be seen as numerical.
• Order: A visual variable is said to be ordered if changes in this visual variable
support ordered readings.

7
 Examples
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation Consider two visual variables:

Data • Position:
Visualization
Gallery • Size:

8
 Data Variables
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation

Data
Type of variable Examples
Visualization
Gallery continuous 1.3, 5.7, 83, 1.5 × 10−2
discrete 1, 2, 3, 4
categorical unordered dog, cat, fish
categorical ordered good, fair, poor
Date or time Jan. 5 2018, 8:03am
Text The quick brown fox jumps over the lazy dog.

9
 Mapping Data Variables onto Visual Variables
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation Concept 3
Data
Visualization Mapping or encoding is a process that converts data values in a systematic and
Gallery
logical way into the visual elements that make up the final graphic.

• A 1-to-1 visual mapping is the standard approach to visual encoding: A single

data variable is mapped onto a single visual variable.
• A 1-to-n visual mapping takes advantage of the combined power of n visual
variables for the encoding a single data variable.

10
 Mapping Data Variables onto Visual Variables (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • To map data values onto aesthetics, we need to specify which data values
Data
Visualization
correspond to which specific aesthetics values.
Gallery • This mapping between data values and aesthetics values is created via scales,
a scale must be one-to-one.

11
 Example
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation (a) Bar chart encode two attributes using a line mark with the vertical
Data spatial position channel for the quantitative attribute, and the
Visualization
Gallery horizontal spatial position channel for the categorical attribute.
(b) Scatterplot encode two quantitative attributes using point marks
and both vertical and horizontal spatial position.
(c) A third categorical attribute is encoded by adding color to the
scatterplot.
(d) Adding the visual channel of size encodes a fourth quantitative
attribute as well.

12
Expressiveness and Effectiveness
 Expressiveness And Effectiveness
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation Two principles guide the use of visual channels in visual encoding:
Data • Expressiveness principle dictates that the visual encoding should express all
Visualization
Gallery of, and only, the information in the dataset attributes.
• match channel and data characteristics
• Effectiveness principle dictates that the importance of the attribute should
match the salience of the channel; that is, its noticeability.
• encode most important attributes with highest ranked channels

14
 Effectiveness Ranking Of Visual Variables
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation
Quantitative Data Ordinal Data Nominal Data
Data
Visualization Position Position Position
Gallery
Length Lightness Shape
Angle Saturation Hue
Area Hue Lightness
Lightness Length Saturation
Saturation Angle Length
Hue Area Angle
Shape Shape Area

15
Evaluation
 Accuracy
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • The obvious way to quantify effectiveness is accuracy.

Data
Visualization
• We perceive different visual channels with different levels of accuracy.
Gallery

17
 Accuracy: Vis experiments
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation Error rates across visual channels, with recent crowdsourced results replicating and
Data extending seminal work from Cleveland and McGill
Visualization
Gallery Cleveland & McGill’s Results

Positions
1.0 1.5 2.0 2.5 3.0
Log Error

Crowdsourced Results

Angles

Circular
areas

Rectangular
areas
(aligned or in a
treemap)

18
1.0 1.5 2.0 2.5 3.0
Log Error
 Discriminability
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • Linewidth has a limited number of discriminable bins.

Data
Visualization
Gallery

19
 Separability vs. Integrality
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • We cannot treat all visual channels as completely independent from each
Data
Visualization
other, because some have dependencies and interactions with others
Gallery

Position Size Width Red

Hue (Color) Hue (Color) Height Green

Fully separable Some interference Some/significant Major interference

interference

20
 Popout
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation Many visual channels provide visual popout, where a distinct item stands out
Data from many others immediately
Visualization
Gallery • Task: Find the red dot
• Question: how long does it take?
• Parallel processing on many individual channels
• speed independent of distractor count
• speed depends on channel and amount of difference from distractors
• Serial search for (almost all) combinations
• speed depends on number of distractors

21
 Popout (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • Visual popout. (a) The red circle pops out from a small set of blue circles.
Data
Visualization
(b) The red circle pops out from a large set of blue circles just as quickly. (c)
Gallery The red circle also pops out from a small set of square shapes, although a bit
slower than with color. (d) The red circle also pops out of a large set of red
squares. (e) The red circle does not take long to find from a small set of
mixed shapes and colors. (f) The red circle does not pop out from a large set
of red squares and blue circles, and it can only be found by searching one by
one through all the objects

22
 Popout (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • Many channels support visual popout

Data
Visualization
Gallery

23
 Example
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • How many 9s are there?

Data
Visualization
Gallery

24
 Example (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • It’s easy to count the 9s.

Data
Visualization
Gallery

25
 Example (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation
• Can you see the 9?
Data
Visualization
Gallery

26
 Example (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • Differences in size are easy to see too.

Data
Visualization
Gallery

27
 Example (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation • Sorted bar chart using color and length to show how many 9s are in our table.
Data
Visualization
Gallery

28
Data Visualization Gallery
 Star Trek: The Next Generation
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation

Data
Visualization
Gallery

30
 A terrible record
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation

Data
Visualization
Gallery

31
 A terrible record (cont.)
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation

Data
Visualization
Gallery

32
 Deaths from malaria, 2000–2014
Marks And
Visual Variables

Expressiveness
and
Effectiveness

Evaluation

Data
Visualization
Gallery

33
References

Goodfellow, I., Bengio, Y., and Courville, A. (2016).

Deep learning.
MIT press.
Munzner, T. (2014).
Visualization analysis and design.
CRC press.
Russell, S. and Norvig, P. (2016).
Artificial intelligence: a modern approach.
Pearson Education Limited.
Ward, M. O., Grinstein, G., and Keim, D. (2015).
Interactive data visualization: foundations, techniques, and applications.
CRC Press.