Reading 20: Too Much Stuff

by Mike Gleicher on April 9, 2015

Due Date: Initial Posting due on Monday, April 13, 11:59pm (ahead of class Tuesday, April 14)

Readings: Munzner 13&14, Clutter or Aggregation Survey (838 only), optional Splatterplots

Discussion: Canvas link

For the initial posting, read Chapter 13 of Munzner and Section 15.3 (about Scagnostics – which we have mentioned in class).

Read Chapter 14 of Munzner as well (you can do this one by Thursday). It’s more about navigation, but that’s also somewhat related to dealing with scale.

These chapters are short. And to be honest, they are not the strongest part of the book.

Unfortunately, the chapter really glosses over dimensionality reduction. And I can’t find a good reference. We’ll talk about it in lecture, many of you have probably seen it already (if you’ve taken an AI class, or a machine learning class, or a multi-variate statistics class). There will be a Vis-Specific reading as an 838-only thing coming up.

For item reduction and aggregation, there are two great survey papers, that help organize all the different kinds of techniques:

If you are an 838 student, you need to read one one these surveys (after Tuesday).  Note: this is not the “838 assignment”, its just you have to do more for this reading assignment.

For the postings:

  1. For your initial posting (before class on Tuesday), discuss how the kinds of techniques that Munzner describes compare to what we’ve seen so far in class. Think of examples we’ve seen (outside of this chapter) that are examples of the techniques she describes.
  2. If you’re an 838 student, describe the basic idea behind the categorization you have read about so that the 638 students in your group (or the 838 students who read the other one) have a sense of what they’ve missed. Give some examples (preferably from class) of the different
  3. In the discussion, try to come up with more examples from class (or that you’ve seen in prior seek and finds, or anywhere else) and see how they are examples of the general approaches from the readings.

In class, we’ll discuss our Splatterplots work (and Alper may even show off his web-based implementation). The paper (which you can find here) is optional.

Print Friendly, PDF & Email

Previous post:

Next post: