Click here to zoom in and actually see this thing.

You need to click on this to see the full size image.

A bit ago I put together a script (details on how to do this in a later post) to keep track of the number of e-mails in my inbox. I keep e-mails in my inbox in Mail.app until I have resolved them (responded, read a paper, performed an action, etc.) keeping them as a bit of a to-do list.

I am really interested right now in passive or ambient indicators. This is an application. Especially because I believe that the amount of e-mail in my inbox is a good indicator of my current busyness, or generally how hectic I am at any given time. I plan to make this a bit better, put it up on the web as more of a sparkline, and also probably a more abstracted level of busyness. I am thinking the scale will be something like:

• 0-20 emails — So Relaxed That This Can Never Happen
• 20-50 emails — Doing Good, Time To Find More Work
• 50-110 emails — Average Busyness Insues, If It Can Wait, Let It Wait
• 110-200 emails — Very Hectic. Too Much To Do. Don’t Expect To See Patrick.
• >200 emails — Might Be Dead, Might Wish He Was.

(Note this idea was borrowed from Dave Shea’s Stress-O-Meter on his contact page at BrightCreative. However his functions differently, it is a constantly increasing counter, that he resets when he has free time, the more he remembers to reset it, the less he has to do, is the model behind it.)

Each year the IEEE holds VisWeek 2008, a series of three co-located conferences: Visualization, InfoVis, and VAST (Visual Analytics in Science & Technology), this year in Columbus, Ohio; October 19th – 24th. Also each year all three of these conferences host visualization conferences, which I of course always think about entering but then never actually do.

Except this year. This past spring Peter Landwehr started up a reading group on Large Scale Visualizations, which slowly shifted – I expect mostly because of me, to a group more on information visualization generally, and at some point we decided to form a group to create an entry for the VAST contest (more on this later). And at the same time as that progressed I convinced Danny that he wanted to spend some time on the InfoVis contest.

The InfoVis contest this year (details here,) focused around over a year’s worth of data from the Mitsubishi Electric Research Lab. They released all this data to the public to allow them to mine and visualize whatever they so chose; and the dataset was also chosen for the 2008 contest.

Danny and I eventually got ourselves together and over the last few weeks put together the following poster. (For a higher resolution version, see my portfolio which has been updated additionally this exists on flickr. )

I expect this is a very different direction than most of the people who entered the contest went in, which may or may not be good, and I have no idea how it will be judged. Either way though, for the amount of time we had, I am quite satisfied with our solution. They contest pages infer they are looking for stories from the data and I think as an overall story, to someone who has never seen this dataset, or worked with sensor data before, it is a nice introduction.

The poster and all of its charts and tables were created in some combination of Apple’s iWork suite and Processing ( processing.org ), the data was mined and refined using some Python and mostly Java.

Now though, the information visualization time of the year is over and it is time to get back to real work. Cell phones and privacy and student organization websites and swappable policy interfaces and paper reading and user testing and rule specification interfaces and other secret and exciting things.