Friday, May 13, 2011

CHI 2011 Top Ten

Just got back from CHI, the biggest Human-Computer Interaction conference. Blew me away. Top-notch research going on from all corners of everywhere. My current group, the Ubicomp Lab at UW, contributed four amazing papers (SkintennaInGenHeatWaveHaptic Laser), but here's the stuff that I saw from other labs that really knocked my socks off:

Personal Data Management
Overlay streams of data on top of each other. Annotate with a special pen. Zoom through video too. This system does it all. I'd be excited to try this out with some lifelogging data.

"What was that guy's address? Ugh, I typed it somewhere, but I don't remember what file it was in, or anything else that was in it, but I was listening to this song at the time." YouPivot helps you find it.

Visualization system for series of events over time. Example: a hospital. Every patient enters, then gets transferred to ICU, Emergency, or Floor, might get shuffled between them, and then leaves or dies. Visualizing one patient's sequence is easy; LifeFlow helps you visualize all the patients' sequences at once.

Eye tracking, heart rate, skin conductivity, and muscle sensors, a temperature monitor near your mouth, a regular game control, and an awesome looking yeti = Death Trigger. They pulled out some principles for designing such multimodal games. Also, looks fun!

Everyone wants to make games that teach, but nobody knows how. These guys actually did a survey of existing game literature, found that almost nobody evaluates effectiveness, and proposed Applied Behavior Analysis as a framework to actually make educational games that work. Super yes! This may have been the most inspirational thing I saw at CHI.

This was just fun. Games in which you punch a shadow opponent on a wall, play table tennis with 3 people, or hang on a bar. Maybe with their help, Dave and Buster's could actually be fun. (or games-for-exercise could actually work.)

Other Stuff
Recording emotion with words is time consuming and introduces another level of errors. (nobody wants to say "I'm depressed.") So these guys made a pretty thorough set of images and phone app to let you record your emotion by picking a photo.

Programmers know that life is better once you switch from mouse menus to keyboard shortcuts. But it's a hard switch to make. These guys propose to help people switch with "Blur", which is basically the Chrome omnibox (address bar) for Windows.

The only theory/design paper to make my list, this made me think about something I hadn't been thinking about: HCI moves super fast, the medical world moves super slow, so how do you make a paper involving a year-long study and get it published? Answer: skip the year-long study. HCI shouldn't necessarily be concerned about that.

Quick way to learn 1000 French words? Replace their English equivalents whenever you're browsing the web. Sure, there are limitations, but language learning seems to me to be a quick burst of getting some basic fluency and grammar followed by a long slog of learning a bunch of nouns/verbs/adjectives. Maybe this can help you skip that.

No comments:

Post a Comment