Foundations%20of%20Visual%20Analytics%20Pat%20Hanrahan%20Director,%20RVAC%20Stanford%20University

1
Foundations of Visual AnalyticsPat
HanrahanDirector, RVACStanford University
2
Analytical ReasoningFacilitated byInteractive
Visualization
3
Why is a Picture (Sometimes) Worth 10,000 Words
4
Lets Solve a ProblemNumber ScrabbleHerb Simon
5
Number Scrabble

• Goal Pick three numbers that sum to 15

6
Number Scrabble

• Goal Pick three numbers that sum to 15
• A
• B

7
Number Scrabble

• Goal Pick three numbers that sum to 15
• A
• B

8
Number Scrabble

• Goal Pick three numbers that sum to 15
• A
• B

9
Number Scrabble

• Goal Pick three numbers that sum to 15
• A
• B

10
Number Scrabble

• Goal Pick three numbers that sum to 15
• A
• B

11
Number Scrabble

• Goal Pick three numbers that sum to 15
• A
• B

?
12
Tic-Tac-Toe
13
Tic-Tac-Toe
X
14
Tic-Tac-Toe
X
O
15
Tic-Tac-Toe
X
X
O
16
Tic-Tac-Toe
X
X
O
O
17
Tic-Tac-Toe
X
X
O
X
O
18
Tic-Tac-Toe
X
X
O
X
O
O
19
Problem Isomorph
3
4
8
5
9
1
7
2
6
Magic Square All rows, columns, diagonals sum to
15
20
Switching to a Visual Representation
8
3
4
5
9
1
7
2
6
21
Switching to a Visual Representation
8
3
4
5
9
1
7
2
6
22
Switching to a Visual Representation
3
4
8
5
9
1
7
2
6
23
Switching to a Visual Representation
3
4
8
5
9
1
7
2
6
24
Switching to a Visual Representation
3
4
8
5
9
1
7
2
6
?
25
Switching to a Visual Representation
3
4
8
5
9
1
7
2
6
26
Why is a Picture Worth 10,000 Words?

• Reduce search time
• Pre-attentive (constant-time) search process
• Spatially-indexed patterns store the facts
• Reduce memory load
• Working memory is limited
• Store information in the diagram
• Allow perceptual inference
• Map inference to pattern finding
• Larkin and Simon, Why is a diagram (sometimes)
  worth 10,000
• words, Cognitive Science, 1987

27
The Value of Visualization

• It is possible to improve human performance by
  1001
• Faster solution
• Fewer errors
• Better comprehension
• The best representation depends on the problem

28
Number RepresentationsNorman and Zhang
29
Number Representations

• Counting Tallying
• Adding Roman numerals
• Multiplication Arabic number systems

XXIII XII XXXIIIII XXXV
30
Zhang and Norman, The Representations of
Numbers, Cognition, 57, 271-295, 1996
31
Distributed Cognition
External (E) vs. Internal (I) process
Roman
Arabic

1. Separate power base I E
2. Get base value E I
3. Multiply base values I I
4. Get power values I E
5. Add power values I E
6. Combine base power I E
7. Add results I E

Arabic more efficient than Roman
32
Long-Hand Multiplication
34 x 72
68 238
2448
From Introduction to Information Visualization,
Card, Schneiderman, Mackinlay
33
Power of Representations

• The representational effect
• Different representations have different
  cost-structures / running times
• Distributed cognition
• Internal representations (mental models)
• External representations (cognitive artifacts)
• Representations 101
• Representations are not the real thing
• Manipulate symbols to perform useful work

34
(No Transcript)
35
Modeling and Simulation

• Simulation for computer graphics is sophisticated
• Diversity of phenomenon
• Complexity of the environment
• Robustness
• Range of models fast to accurate
• Lots of breakthroughs one small example is GPUs
  which may become the major platform for
  scientific computation

36
Mathematics of Visual Analysis

• MSRI, Berkeley, CA, Oct 16-17, 2006
• Organizers P. Hanrahan, W. Cleveland, S.
  Harabagliu, P. Jones, L. Wilkinson
• Participants J. Arvo, A. Braverman, J. Byrnes,
  E. Candes, D. Carr, S. Chan, N. Chinchor, N.
  Coehlo, V. de Silva, L. Edlefsen, R. Gentleman,
  G. Lebanon, J. Lewis, J. Mackinlay, M. Mahoney,
  R. May, N. Meinshausen, F. Meyer, M.
  Muthukrishnan, D. Nolan, J-M. Pomarede, C. Posse,
  E. Purdom, D. Purdy, L. Rosenblum, N. Saito, M.
  Sips, D. W. Temple Lang, J. Thomas, D.
  Vainsencher, A. Vasilescu, S. Venkatasubramanian,
  Y. Wang, C. Wickham, R. Wong Kew

37
Supporting Interaction

• Panelists William Cleveland, Robert Gentleman,
  Muthu Muthukrishnan, Suresh Venkatasubramanian,
  Emmanuel Candez
• Fast algorithms streaming and approximate
  algorithms, compressed sensing, randomized
  numerical linear algebra,
• Fast systems map-reduce, column stores, beyond
  R,

38
Finding Patterns

• Panelists Peter Jones, Vin de Silva, Francois
  Meyer, Naoki Saito, Michael Mahoney
• How to represent patterns?
• Data/dimensional reduction vs. transformation to
  meaningful form?
• Are humans required to build good models? How is
  domain knowledge added?
• When are computers good pattern finders? When are
  people good pattern finders?

39
Computation Steeringvs.Interactive Simulation
40
Integrating Heterogenous Data

• Panelists Sanda Harabagliu, John Byrnes,
  Jean-Michel Pomeranz, Christian Posse, Guy
  Lebanon
• Many important datatypes text and language,
  audio, video, image, sensors, logs, transactions,
  nD relations,
• How to fuse into common semantic representation?
• Beyond the desktop to new representations of
  information spaces vispedia, jigsaw,

41
Smart Visual Analysis

• Panelists Leland Wilkinson, Jock Mackinlay, Jim
  Arvo, Amy Braverman, Dan Carr
• Automatic graphical presentation and
  summarization guided analysis
• How do people reason about uncertainty?

42
Summary

• Visual analytics merges
• Cognitive psychology
• Mathematics and computation (algm, stat, nlp)
• Interactive visualization techniques
• Need to rethink how these capabilities are
  combined