The core premise of this project is for you to develop a visualization which offers insights into some data you care about.
- This assignment (#30) is due at the beginning of class on Tuesday March 4.
- Prior to this, a Looking Outwards on the topic of Interaction is due on February 18th.
- By 2/20 you should also have your data in hand, or have a method for obtaining it.
- Immediately after, a Looking Outwards for your capstone project is due March 6th.
Upon completion of this assignment, students will be able to demonstrate familiarity with the complete pipeline of operations in information visualization, from data acquisition, parsing, and cleaning to representation, distillation and interaction.
For many or most of you, this project will probably take the form of a “Quantified Selfie” — a computational self-portrait developed from any (one or more) of the data-streams you produce. Possible data sources include your personal fitness trackers, such as the Nike Fuelband and/or Fitbit One — as well as your other electronic communications and information trails, such as your Gmail, Twitter stream, Facebook wall, SMS messages, phone call logs, GPS movements, financial transactions, etcetera. You may also compare, contrast and correlate these data with “external” sources of information, such as (for example) world news, weather, or the information patterns produced by your friends and family.
NOTE 0. Operate from the standpoint that your visualization is a tool you’re building to help you answer a question. What are you curious about? (Why?) Can you collect data about phenomena that nobody has ever seen or thought of before?
NOTE 1. You are encouraged (but not strictly required) to combine multiple sources of data, in order to establish frameworks within which interesting comparisons can be made, and unexpected insights can be drawn. Think outside the box: What if you somehow contrast your text messages with the collected work of William Shakespeare (5MB)? etc…
NOTE 2. You are encouraged (but not required) to create an interactive (as opposed to a static) visualization. Consider how your project can support (one or more of) the core interactive operations of zooming, sorting, filtering and/or querying. Think about how your user can explore your data, in order to experience a large data set in detail. Is interactivity required? Well, not strictly; it is acceptable to generate a high-resolution PDF if that presentation format suited your concept well — or even, a 3D printed sculpture, if you have access to that kind of output. Likewise, I know that some of you are very musically oriented — you are welcome to consider making a information sonification.
NOTE 3. Not everything is a timeline, just because you happen to have data with a timestamp. Consider your email, for example. While you can certainly visualize this as a timeline (e.g. the number of emails you sent and received per day), you could also view it as a graph (e.g. the social network of your friends and acquaintances), or as a histogram (e.g. which people you communicate with most, or which words you use most often).
NOTE 4. You don’t haaaave to make a self-portrait. If there is some data you would prefer to explore, focus on that. In this class, you should always focus on the subject which captures your imagination and curiosity. The ‘Quantified Selfie’ premise (and the fitness trackers) are offered as a starting point.
NOTE 5. Give yourself permission to be specific. Sometimes focusing your investigation on a subset of your data is not only simpler, but can be much more conceptually interesting, than trying to visualize the “whole thing”. For example, instead of visualizing all of your text messages (a study of your texting behavior), what if you only examine the ones you exchanged with a certain person (a study of your relationship)? Another way to think about this is: give yourself permission to have a point of view.
Readings & Background
- Maureen O’Connor. “Heartbreak and the Quantified Selfie“. NY Magazine, 12/2/13.
Abstract: Quantified Selfies offer the chance for self-discovery. Because digital data is vast and immaterial, viewing the aggregate can be difficult, and the process of organizing information can reveal unexpected truths.
- Robert Crease, “Measurement and Its Discontents“. New York Times, October 22, 2011.
Abstract: Why are we still stymied when trying to measure intelligence, schools, welfare and happiness? The problem is not that we don’t yet have precise enough tools for measuring such things; it’s that there are two wholly different ways of measuring.
- Judith Donath et al. “Data portraits“. Proceedings of SIGGRAPH 2010, Pages 375-383.
Abstract: Data portraits depict their subjects’ accumulated data rather than their faces. They can be visualizations of discussion contributions, browsing histories, social networks, travel patterns, etc. They are subjective renderings that mediate between the artist’s vision, the subject’s self-presentation, and the audience’s interest. Designed to evocatively depict an individual, a data portrait can be a decorative object or be used as an avatar, one’s information body for an online space. Data portraits raise questions about privacy, control, aesthetics, and social cognition. These questions become increasingly important as more of our interactions occur online, where we exist as data, not bodies.
Finally, be sure to look at these resources for your project:
- IACD Quantified Selfie resources
- Some information visualization things to view
- Tools which may help you collect data about yourself, including The Quantified Self Guide, the Beginner’s Guide to Quantified Self, and other tools like:
Nick Felton’s new Reporter app
- Tools which may help you collect data about the outside world, such as Temboo, which provides interfaces from 7 languages to more than 100 API’s; or the KimonoLabs tool.
- Tools which may help you pre-visualize your work, such as the IBM ManyEyes system.
- Nike Fuelband data scraper and parser (by Wes Grubbs & Golan Levin)
- Temboo-based Fitbit data scraper
- Create a blog post to host your project.
- If it is possible to embed your visualization in the blog post directly (e.g. D3, PDF, ProcessingJS, etc.), please do so. Contact the professor if you need any special WordPress plugins installed.
- Document your visualization in a video, using screen-capture software — even if it is possible to embed your visualization. It’s helpful if your video has narration or titles. Upload this video to YouTube or Vimeo, and embed this in your blog post.
- At the top of your blog post, please write a tweetable one-sentence overview of your project.
- In approximately 200-300 words, describe your project. Consider: What data did you use, and why? What inspirations did you have (or what prior work did you research?) What did you learn in making it? What design choices did you make? What would you do better if you could do it again, or if you had more time?
- What does your visualization reveal about your data? Discuss some general patterns you discovered in your data, and discuss some specific and unique datapoints that stand out as outliers. In the blog post, include a few screenshot images which illustrate these patterns and callouts.
- Feel free to include scans of your preliminary sketches, and/or images citing your inspirations, as well.
- Include a link to your Github repository where your code for the project is kept.
- Categorize your blog post, 30-Visualization.