CS295J/Theory of Visualization outline: Difference between revisions

From VrlWiki
Jump to navigation Jump to search
New page: there are different kinds of models, but permit me to be somewhatinformal here. CHARACTERISTICS OF A USEFUL THEORY OR MODEL *explains observations *simplifies -- doesn't just repeat obser...
 
 
(One intermediate revision by the same user not shown)
Line 21: Line 21:
*[VAST keynote speaker said: for vis tools, utility is speed of doing something.
*[VAST keynote speaker said: for vis tools, utility is speed of doing something.
*how quickly they can be used
*how quickly they can be used
*create models/theories to simplify/predict human
*create models/theories to simplify/predict human
**speed/accuracy using it for important tasks
**speed/accuracy using it for important tasks
Line 42: Line 41:
We actually have some ingredients
We actually have some ingredients


some visualization methods range from
some visualization methods; range from
*often used to
*often used to
*invented but not used
*invented but not used
*they give us some sense of the shape of the space of visualization methods
*they give us some sense of the shape of the space of visualization methods


some jobs or tasks
some jobs or tasks; they range from
*range from
*the abstract ``analyze data''
**the abstract ``analyze data''
*to ``count the critical points in this picture of a 2D vector field''
**to ``count the critical points in this picture of a 2D vector field''


some acquired data or observations about these visualization methods and tasks
some acquired data or observations about these visualization methods and tasks

Latest revision as of 18:22, 23 January 2009

there are different kinds of models, but permit me to be somewhatinformal here.

CHARACTERISTICS OF A USEFUL THEORY OR MODEL

  • explains observations
  • simplifies -- doesn't just repeat observations
  • predicts future observations


theories of physics

  • started with lots of observations of the world
  • created models/theories to simplify/predict them
  • Newtonian
  • relativistic
  • quantum
  • maybe string theory


we'll start a visualization theory with observations about visualization methods

  • we observe how useful they are
  • [VAST keynote speaker said: for vis tools, utility is speed of doing something.
  • how quickly they can be used
  • create models/theories to simplify/predict human
    • speed/accuracy using it for important tasks

u(v,j)

  • u tells us how good a visualization method is for a task
  • space of all possible visualization methods
    • and all possible tasks
  • we just have to figure out how to calculate it

oh, and ``visualization method

  • all possible interactions
  • all possible display setups

this is very meta-level

  • but it gives us a starting point
  • at least for a thought experiment

We actually have some ingredients

some visualization methods; range from

  • often used to
  • invented but not used
  • they give us some sense of the shape of the space of visualization methods

some jobs or tasks; they range from

  • the abstract ``analyze data
  • to ``count the critical points in this picture of a 2D vector field

some acquired data or observations about these visualization methods and tasks

  • these data range from
    • unrecorded experiences and intuition
    • through recorded anecdotal data
    • to quantitative user study results
  • these data are effectively point samples of u
  • some can perhaps be generalized
  • but we're not quite done yet

some models/theories of interaction tell us how we might calculate sub-parts of u*

  • Fitt's Law
  • Steering Law
  • predictions of how parameterized interaction techniques work
    • how accurate/fast will people be

a pretty good understanding of visual perception

  • models of low-level vision: center/surround of photo receptors
  • models of cue integration: how we couple and transition among
    • the handful of different depth cues, for example
  • models of higher-level vision: sensory cues used for navigation
  • related to visualization and interaction, but aimed at how people make sense of their senses, not how vis techniques work
    • [sounds subtle, but for now it's a significant difference]

theories of cognition

  • e.g. visual working memory
  • e.g. verbal working memory

finally, visual design, product design, art

  • best practices
  • tailored over centuries
  • experience that is taught through doing
  • and that very much incorporates evaluation