What is InfoVis?
It is a process of transforming information into a visual form enabling the viewer to observe, browse, make sense, and understand the information.
In InfoVis what is shown?
Data Abstraction
In InfoVis why is the user looking at it?
Task Abstraction
In InfoVis how is it shown?
Idiom: visual encoding and interaction
What can be visualized?
To answer this questions:
What dataset?
What types of data?
Why people are using vis in terms of actions and targets?
To answer this questions:
What tasks?
What purpose?
How to design vis idioms: encode, manipulate, facet, and reduce?
Techniques used to visually encode data.
What are the four levels of Design in InfoVis?
- Domain situation
- Data/task abstraction
- Visual encoding/interaction idiom
- Algorithm
What is the domain level?
The users
Their questions
The vocabulary
The data
What is the abstraction level?
Abstract into domain independent vocabulary.
Appears different in the problem domain, actually same task.
What is the idiom level?
Create the visual representation.
What is the algorithm level?
How to actually get it done!
What are the two “Angles of Attack”?
- Problem-driven
- Technique-driven
What is Problem-driven?
Top-Down
Have a question, want to have a good way of answering it.
What is Technique-driven?
Bottom-Up
Want to develop a new algorithm.
No idiom seems to work for the tasks at hand.
What are the five basic data entities?
- Items
- Attributes
- Links
- Positions
- Grids
What is the first aspect to consider when designing a problem-driven visualization?
With problem-driven work, you start at the top domain situation level and work your way down through abstraction, idiom, and algorithm decisions.
There are four different types of search actions in information visualization. What are they and when do they take place, depending on the nature of data?
- Lookup: find it where you expect it.
ex: locate yourself on your family tree. - Browse: know where it is, but not what it is.
ex: who were my great-grandfather’s brother’s sons? - Locate: I know what, but not where.
ex: how is Paul related to me in the family tree? - Explore: look for interesting things.
ex: who were my ancestors? How did they relate to one another?
What is the best channel to encode a quantitative variable: position or angle? Why?
Position.
The attributes encoded with position will dominate the user’s mental model — their internal mental representation used for thinking and reasoning — compared with those encoded with any other visual channel.
Color consists of different components. Color hue, in particular, is more correctly used to encode attributes of a particular type. Name this attribute type and discuss the reason why this rule must be followed.
Identity Channels: Categorical Attributes.
The major design choice for color-map construction is whether the intent is to distinguish between categorical attributes or to encode ordered attributes.
John needs to create a visualization to use in his neuroscience research to study brain cortex areas and their functions in mice. He asked his friends for help but they don’t seem to agree. While Mary argues that 3D should be avoided at all costs when creating an effective visualization, Joseph believes that it is the best option in this case. With whom do you agree and why?
If we have a need for visualization in clinical context and in a situation where we need to analyze three dimensional models, the 3D option is certainly better as it will also have to be production information in this case (eg: cuts, slices, etc…).
What is the Pre-Attentive Processing (“Popout”)?
Low-level tasks that can be accomplished almost instantaneously (<250msec) without focused attention. Objects seem to “pop-out”.
What is a mark?
The visual element (line, star, bar, wedge, etc.).
What is a channel?
The visual property (length, size, shape, etc.).
