Spring 2007 - Assignments

These assignments are designed to challenge you to think about information visualization, expose you to some existing tools, and to try your hand at developing visualizations.

Some of these assignments may require downloading and installing of software. Some tools will also require installing the Java Runtime environment. If you do not have an access to a PC where you can install software, or if you run into any problems, please let me know.

All assignments should be submitted electronically, and should be handed in by 4pm on the due date, so that materials can be incorporated into the lecture when appropriate.

• Assignment 1, Due Feb. 6
• Assignment 2, Due Feb. 27
• Assignment 3, Due April 10
• Assignment 4, Due April 24

Some of these assignments ask you to provide screen shots illustrating the use of various visualization tools. There are shareware and freeware tools available for Mac, Linux, and OS X that can be used for this purpose. I personally use the screen capture facilities in the GIMP software, which also supports editing, resizing, and retouching of images.

Please keep in mind that the class project has additional due dates: the proposal is due on March 6, the progress report is due on April 17, presentations are on May 10 and 15, and the final report is due on May 17.

Assignment 1: Static Visualizations Due Feb. 6

Find two examples of data visualization. The examples can be charts, graphs, networks, or any other representation. One of these should be an example of data visualization done well, while the other should be an example of visualization done poorly.

Possible sources include (but are not limited to) web sites, newspapers, magazines, books, maps, and advertisements.

For each example, provide the following:

• The source (URL, newspaper/magazine citation, etc
• A picture (electronically, if possible).
• A short (at most one page) description of why this is an effective/ineffective visualization.

Be prepared to discuss and defend your choices in class.

Assignment 2: Hierarchical Visualization Due Feb. 27

The goal of this assignment is to gain some experience using interactive information visualization tools.

Compare and contrast two tools that use different strategies to visualize hierarchical information. For each tool, use the visualization to interpret and explore one or more datasets.

Tools that you might consider using include:

• Treemaps: SequoiaView or Treemap 4.1. These tools will both let you visualize the contents of your hard drive - a great test case.
• Hyperbolic Browsers: SiteLens or StarTree . (Note that these tools will not allow you to load your own data - which is ok.)
• Other node-link viewers, such as SpaceTree or Stretch

You can use the data samples provided by the tools (or websites),or your own data.

You may use other tools if you find any, but the two tools should not use the same basic strategy. In other words, don't use both SequioaView and Treemap, as those are both treemap based.

Write a short (at most 5 pages) review and critique of these tools, discussing the following:

• The data sets used
• Any insights or understanding that you gained from the tools
• Comparative strengths and weaknesses of the tools
• Changes that you might make to improve the tools.
• Changes that you might make to improve the tools.
• Which types of data sets might either tool be best used for?

In addition, please provide one or two screen-shots of the tools in action.These screen shots will not be included in the 5-page limit for the writeup.

Please send the report to me by email before 4pm on Feb. 27. Be prepared to discuss and defend your choices in class.

Assignment 3: Multivariate Data Due April 10

This assignment is similar to assignment 2. Find a multivariate data set on the web, using one of the suggested data sources or from any other source, and use a multivariate visualization tool to explore and interpret it . Some tools are given on the links page. Others may be found by searching the web.

Assignment 4: Visualization Histories Due April 24

The retrieval interesting and useful insights is an important part of the use of information visualization for meaningful analytics. Unfortunately, the definition of what is "interesting" evolves over time. This leads to the very real possibility that items identified in a visualization may not be deemed interesting until long after the user has moved on to a different view of the data.

One approach to handling this problem is to store all user actions, leading to a set of visualization histories, tracking the state of the visualization tool at different points in time. These histories can be replayed or reviewed to identify items that may have been seen earlier. As the goal of such systems is to ease the process of recovering lost insight, these visualizations must be designed to accentuate cues that might help users find data of interest.

In this assignment, you will design a history visualization mechanism for one of the tools that we have examined in class. You will choose a tool, design a visualization for interaction histories, enumerate the data items and types that these histories would need to store, and describe how a user might interact with it.

Your 3-6 page writeup should include your summarization of the problem, description of your proposed design, justification of the design decisions, and any other details that might be useful for convincing someone to adopt your ideas. Interface "mock-ups" - in the form of pictures or other graphics that illustrate your ideas - can be very useful, but you should not put huge amounts of time into developing professional-quality prototypes: simple displays that convey the relevant information are sufficient.

---