Due Date: Friday, February 20th, 11:59p
Update (2/18): in retrospect, this assignment was a bit much. You didn’t get a full 2 weeks, and the overlap with previous readings may not be as helpful as I thought. So, I am reducing the assignment. The recommendation is now 4 papers (you pick which 1 of the 5 to skip). If you do read more than 4, let us know. Also, it is probably better to read 4 papers quickly than 3 papers carefully. Part of the lesson here is to learn to read things quickly to get the important bits (a valuable skill with papers). In fact, skimming all of the papers to get the gist of them is probably better than even reading 5 more carefully.
I was going to make this Monday, the 16th (since I said things would be due on Mondays), but wanted to give people more time. In the future, I think 838 assignments will be due on fridays, so things don’t pile up too much.
Turn-in link: 838-Only: Original Sources
Note: this is assignment is required for CS838 students. If you’re a 638 student you’re welcome to do it.
In class, we’re primarily reading the textbook (so far) for some of the basic stuff. This has the down side that you’re getting one perspective (the textbook author’s), and you miss out on some of the historical development.
For this assignment, I’d like you to read some (yes, not all of the below!) of the papers that lead to the early chapters in Munzner’s book, to get a broader perspective on some of the topics we’ve spoken about.
Task Taxonomies (read 1 required, pick 1):
- Munzner’s Chapter 3 is a distilled version of her InfoVis 2013 paper “A Multi-Level Typology of Abstract Visualization Tasks” with her student, Mathew Brehmer. (you can find it here). Most of the main ideas of the paper you’ve seen in the Chapter. The main thing the paper adds is extensive references to a lot of previous work on understanding tasks, and some comparison amongst them.
- (required) Concurrent with that paper was a paper “A Design Space of Visualization Tasks” which takes a different tact (and is also a little more realistic that it doesn’t cover everything). The paper is here.
- While there are many old taxonomy papers, Schneiderman’s “The Eyes Have It” paper from 1996 (here) is one of the most influential. He has a very specific idea about how people were going to use visualization, and almost seems to suggest that this is the only way.
- There were a series of papers by Robert Amar and John Stasko (and some of Stasko’s other students). These stand out to me because (1) it won a best paper award at the first InfoVis conference I went to, and (2) while they are focused in terms of the topic they consider (decision making), they were pretty deeply thinking in terms of the problems they saw people tackling. The Paper “A Knowledge Task-Based Framework for Design and Evaluation of Information Visualizations” was InfoVis 2004 (best paper winner). “Low-level components of analytic activity in information visualization” (infovis 2005) is a little less memorable.
Evaluation:
Munzner’s Chapter 4 is a reframing of her paper “A Nested Model for Visualization Design and Validation.” Infovis 2009 (project page with pdf) The basic ideas are the same, although it is presented with more emphasis on evaluation.
Why Visualize:
I left this paper off of the reading list for the class. The person who wrote it is an important person in the community who still is winning best paper awards and coming up with important new ideas. There are many versions of it – it started out as a position paper at a conference, evolved into multiple different extended version, and I think this is one of the more definitive and personal ones. (some of the other papers had other co-authors, and dilluted the message of the author in expressing his points in as sharp and opinionated a way). I am not sure I like the economic model he proposes, but it is food for thought. The historical perspective is interesting.
Jarke J. van Wijk. Views on Visualization. TVCG, 12(4) 2006.
The (formerly) Optional Anti-Tufte
Michelle Borkin et. al. What Makes a Visualization Memorable? [pdf] InfoVis 2013. I chose to have you read the chart-junk paper, since there was good follow-on debate. But this one could have been the required paper instead. We can argue whether this experiment is a good model for looking at read visualizations, but it is an interesting set of observations.
The Assignment
I want you to read 5 things (update 2/18: see note above. you can choose to skip 1 of the 5, although it is best to read more things less closely) (in addition to the regular readings for class). From the Task category, read #2 (Design Space), and one other paper. Then read Munzner’s evaluation paper (will be an easy read since you’ve seen most of it before) and van Wijk. And Borkin (if you haven’t already).
For the discussion, generally comment on what you’ve read (be sure to say which papers you are including). The big question is “what did you get from these beyond the readings we did for class?” Another question you can consider “should these papers have been part of the required reading list?”
Post your response into the discussion on Canvas. As with all reading discussions, please respond to your peers up until a week after the due date.
