An Overview of Exploratory Data Visualization

1
An Overview of Exploratory Data Visualization

• Dr. Matthew Ward
• Computer Science Department
• Worcester Polytechnic Institute

2
What is Visualization?

• Graphical presentation of data and information
  for
• Presentation of data, concepts, relationships
• Confirmation of hypotheses
• Exploration to discover patterns, trends,
  anomalies, structure, associations
• Useful across all areas of science, engineering,
  manufacturing, commerce, education..

3
Visualization Through History

• Hieroglyphics
• Charts
• Maps
• Diagrams

4
Visualization Today

• Medicine
• Earth Sciences
• Life Sciences
• Engineering
• Manufacturing
• Economics/Commerce
• Communications

5
The Visualization Process
Raw Data
Filter, Select
Transform, Aggregate
Derived/Extracted Data
Normalize
Map Data Components
Graphical Components
Reorganize, Sort
Present One or More Ways
Zoom, Rotate
Display
6
Data Characteristics

• Continuous Model (mostly SciVis)
• Number of independent variables (1, 2, 3, n)
• Data type (scalar, vector, tensor, multivariate)
• Number of dependent variables (1, many)
• Discrete Model (mostly InfoVis)
• Connected
• Graphs, trees, node-link, hierarchical
• Unconnected
• Dependent or independent variables (2, 3, n)

7
Graphical Mappings

• Position (x, y, z)
• Color (hue, saturation, value)
• Shape (need to be perceptually distinct)
• Size
• Orientation (can interfere with shape)
• Texture (contrast, orientation, frequency)
• Motion (2 or 3 D)
• Blinking

8
Many Perceptual Issues

• How accurately do we perceive various graphical
  features?
• How quickly can we detect/classify something
  visually?
• How are our abilities affected by training?
• How variable is our perception based on the
  surrounding field of view?
• How is our perception affected by stress, age,
  gender, boredom, fatigue.

9
1-D Techniques
10
2-D Techniques
11
3-D Techniques
12
N-D Techniques
13
Dynamic Techniques
14
Nontraditional Techniques
15
The Need for Interaction

• All stages of the visualization pipeline can
  benefit from user interaction
• Exploration requires tools for navigation,
  filtering, selection, view enhancement
• Much of recent innovation has focused on
  developing intuitive, powerful interaction
  mechanisms
• Interactions can focus on objects, their
  attributes, or their interrelationships

16
Some Interactive Tools
17
Summary

• Visualization is a powerful tool for qualitative
  analysis of data and information
• It can be useful for presenting or exploring
  virtually any data, regardless of size, type,
  complexity, or application domain
• It can be effectively used to detect, isolate,
  and classify data features of interest and guide
  and evaluate the results of quantitative data
  analysis

18
Visualization Resources - Books

• Keller, Peter, and Keller, Mary. Visual Cues
  Practical Data Visualization. IEEE Press, 1993.
• Tufte, Edward. The Visual Display of Quantitative
  Information. Graphics Press, 1983.
• Tufte, Edward. Envisioning Information. Graphics
  Press, 1990.
• Tufte, Edward. Visual Explanations. Graphics
  Press, 1997. .
• Fayyad, Usama, et. al.. Information Visualization
  in Data Mining and Knowledge Discovery.
  Morgan-Kaufmann, 2002.
• Nelson, Gregory, et. al.. Scientific
  Visualization Overviews, Methodologies,
  Techniques. IEEE CS Press, 1997.
• Lichtenbelt, Barthold, et. al. Introduction to
  Volume Rendering. Prentice-Hall, 1998
• Spence, Robert. Information Visualization.
  Addison-Wesley, 2001.
• Ware, Colin. Information Visualization
  Perception for Design. Morgan-Kaufmann, 1999.
• Chen, Chaomei. Information Visualization and
  Virtual Environments. Springer, 1999.

19
Visualization Resources - Journals

• IEEE Transactions on Visualization and Computer
  Graphics
• Information Visualization
• Computer Graphics and Applications
• Journal of Computational and Graphical Statistics

20
Visualization Resources - Conferences

• IEEE Visualization Conference
• IEEE InfoVis and Volume Visualization Symposia
• SPIE Conference on Visualization and Data
  Analysis
• Eurographics Visualization Symposium
• ACM Symposium on Software Visualization
• Int. Symposium on Intelligent Data Analysis
• Int. Conference on Information Visualization
• ACM SIGKDD