Information Visualization Failed Experiment or Future Revolution

1
Information VisualizationFailed Experiment or
Future Revolution?

• Information Architecture Summit 2004February
  27-29, 2004Austin, Texas
• Karl FastUniversity of Western Ontario
• karl.fast_at_pobox.comwww.livingskies.com

2
A simple question with no good answers

• Does anyone have references to articles that
  explore the relationship of information
  architecture with information visualization? My
  searching is unfruitful. Thanks for your help.
• Daniel Drop, Nov 08 2003http//www.info-arch.org
  /lists/sigia-l/0311/0060.html

3
An interesting response

• I've been something of a renowned InfoViz crank
  on this mailing list... When I first came across
  it, it seemed like it had such promisebut then,
  year after year after year, little if any
  progress occurs.
• The gist infoviz doesn't work, by and large,
  because it's just not usable, and it doesn't
  allow people to accomplish tasks in any better
  way.
• Peter Merholz, Mon Nov 10 2003http//www.info-ar
  ch.org/lists/sigia-l/0311/0074.html

4
Skepticism about information visualization

• I'm a true skeptic about the mass market
  potential of information visualization approaches
  to search.
• Im a skeptic, but a very curious and
  open-minded one.
• Peter Morville, Jan 7. 2004,AIfIA members
  mailing list

5
Skepticism on information visualization

• Skepticism is healthy
• Not a universal sentiment
• However, information visualization doesnt have a
  high profile in information architecture

6
Two purposes of graphics
To communicatean idea
To generate, create, anddiscovernew ideas
7
A definition

• Information visualization is the use of
  computer-supported, interactive, visual
  representations of abstract data toamplify
  cognition.

Readings in Information VisualizationStuart
Card, Jock Mackinlay, Ben Shneiderman (1999, p.
7)
8
When was it above 20ºC during this month?
9
Now when was it above 20?
Mon
Sun
15 17 19 15 22 10 15 15 10 11 15 20 12 18 14 23
12 15 18 12 17 13 11 21 10 29 12 17 22 12
10
Now add some interaction controls
Mon
Sun
15 17 19 15 22 10 15 15 10 11 15 20 12 18 14 23
12 15 18 12 17 13 11 21 10 29 12 30 22 12
0
35
20
29
11
Sliders simplify queries and exploration
Mon
Sun
15 17 19 15 22 10 15 15 10 11 15 20 12 18 14 23
12 15 18 12 17 13 11 21 10 29 12 30 22 12
0
35
10
12
12
Lets add more months
15 17 19 15 22 10 15 15 10 11 15 20 12 18 14 23
12 15 18 12 17 13 11 21 10 17 12 30 22 12
15 17 19 15 18 10 15 15 10 11 15 16 12 18 23 23
27 34 32 29 27 13 11 21 10 29 12 30 22 12
15 17 19 15 11 10 15 15 10 11 15 08 12 18 14 15
12 15 18 12 17 13 11 15 10 10 12 10 11 12
0
35
20
29
13
Lets abstract out the numbers
15 17 19 15 22 10 15 15 10 11 15 20 12 18 14 23
12 15 18 12 17 13 11 21 10 17 12 30 22 12
15 17 19 15 18 10 15 15 10 11 15 16 12 18 23 23
27 34 32 29 27 13 11 21 10 29 12 30 22 12
15 17 19 15 11 10 15 15 10 11 15 08 12 18 14 15
12 15 18 12 17 13 11 15 10 10 12 10 11 12
0
35
20
29
14
Now we can add more data
0
35
20
29
15
A real example the DateLens
http//www.cs.umd.edu/hcil/datelens/
16
A real example the DateLens
http//www.cs.umd.edu/hcil/datelens/
17
Some key ideas

• Visual representation of information
• Exploit visual perception
• Visual perception is powerful and fast
• Changes the cost structure of a task
• Shift the cognitive onto the perceptual
• Interaction with this representation to create
  insight and understanding
• Knowledge crystallization

18
Knowledge Crystallization
Readings in Information VisualizationStuart
Card, Jock Mackinlay, Ben Shneiderman (1999, p.
10)
19
The classic role of the author
RawData
Presentation
Selection
Encoding
User
Author
Information VisualizationRobert Spence (2001 p.
13)
20
The NEW role of the author
RawData
Presentation
Selection
Encoding
User
Visualization Tool Designer
Information VisualizationRobert Spence (2001 p.
13)
21
Common Misconceptions

• Information visualization
• Aims to eliminate text
• Will only succeed by solving the scalability
  problem
• Is about doing things faster, like searching
• Is essentially Tufte
• Will take over the world
• Is sufficiently advanced that it is now a
  commercialization problem

22
Pulleys
Structural
Functional
http//www-psych.stanford.edu/pauly/Publications/
2000/diagrams2000_tversky_etal.pdf
23
Emerging research fronts
Morris, S. A., Yen, G., Wu, Z., Asnake, B.
(2003). Time line visualization of research
fronts. Journal of the American Society for
Information Science and Technology, 54(5),
413-422.
24
Classic example The film finder (1993)
25
Political book purchases interaction?
http//orgnet.com/divided.html
26
Authorlink
http//cite.cis.drexel.edu/
27
Interacting with polyhedra
http//publish.uwo.ca/kamrans/
28
4D Polytypes

• Through interaction with these representations,
  different cognitive tasks can be performed, such
  asvisualizing,analyzing,interpreting,meaning
  making, decision making, planning, problem
  solving,and learning.

Sedig, K., Morey, J. (2000) A Descriptive
Framework for Designing Interaction for Visual
Abstractions. http//publish.uwo.ca/kamrans/_VRI
2002_1.pdf
29
Some challenges

• Finding the right form of representation
• Developing meaningful interaction
• Understanding the connections between type of
  interaction and the cognitive task
• Effective evaluation techniques
• Developing a prescriptive framework to guide
  designers
• Intranets, digital libraries, information-heavy
  sitesknowledge environments

30
Back to the skepticism

• I've been something of a renowned InfoViz crank
  on this mailing list... When I first came across
  it, it seemed like it had such promisebut then,
  year after year after year, little if any
  progress occurs.
• The gist infoviz doesn't work, by and large,
  because it's just not usable, and it doesn't
  allow people to accomplish tasks in any better
  way.
• Peter Merholz, Mon Nov 10 2003http//www.info-ar
  ch.org/lists/sigia-l/0311/0074.html

31
Skepticism about information visualization

• I'm a true skeptic about the mass market
  potential of information visualization approaches
  to search.
• Im a skeptic, but a very curious and
  open-minded one.
• Peter Morville, Jan 7. 2004,AIfIA members
  mailing list

32
As we may thinknotAs we may search
33
Its not just about retrieval

• Tendency to focus on retrieval in classic sense
• Are precision, recall, and time on task useful
  measures?
• The sentiment You cant visualize what you cant
  retrieve is misleading because it
• Suggests people arent finding things
• Suggests that we should not look beyond
  retrieval until we perfect retrieval
• Frames the act of finding a document as the
  central step in communicating ideas and
  understanding them, downplaying what people
  actually do with the documents they have found

34
Big themes

• A significant research problem
• Not always applicable to information retrieval as
  we have traditionally conceptualized it
• The same goes for information architecture
• Meaningful interaction to support learning and
  thinking and reasoning
• Not a failed experiment
• But not a future revolution eitheryet

35
As we may thinknotAs we may search

36
Thank you

• Karl FastCognitive Engineering LabFaculty of
  Information Media StudiesUniversity of Western
  OntarioLondon, Ontario
• For more details about this presentationhttp//
  www.livingskies.com/writings/2004/ia-summit/

karl.fast_at_pobox.comwww.livingskies.com