Emerging Tech: June 2006 Archives
Read about GSM Location on 7.5th Floor, one of my 'dailies'. The project looks interesting, as a new viral location-aware tech. One of these days I'll build that Python app...
Lots of cool stuff in the mix over at GumSpots.com. Check it out - the PicPass and 2cents services are super cool, and GSPS (GumSpots positioning system) is well on its way. Big ups to Kaufman for presenting at the Where Conference next week.
Last Thursday, June 1, I attended the UMD Human Computer Interaction Labs Annual Symposium. Students presented a number of really interesting projects on subjects ranging from data-vis to network analysis to interface design. The day was divided into Visual Interfaces, Public Access, and Interaction & Devices.
The presentations opened with a talk by Ben Schneiderman on creativity. I thought the address was pretty dull and academic, though some interesting areas were touched upon. I've always found any kind of studied tool to enhance creativity to be... well... bullshit. But what's an opinion worth...
Another dissapointing talk came soon after, on the topic of search catagorization. I found this obtuse and outmoded, completely uninformed by all the recent positive developments in internet tech and culture. Kules and Schneiderman argued for 'meaningful and stable categories' — which is a hugely silly notion. There was no question of the authority imposing said categories, and how said categories can evolve, morph, change with the context.
James Rose, Catherine Plaisant, Matt Kirschenbaum, and others presented a body of work on data mining and visual interfaces that I found intelligent and enjoyable. The students took the body of Emily Dickenson's letters and studied the erotic content utilizing bayesian filtering. I was dissapointed at the weak metric used for classification—basically the hot-or-not interface—I found it weak. Would have been interesting to see some biometrics at work, or something with more sophistication.
Adam Perer presented a visually intense network-vis project. The rank-by-feature framework seemed a little goofy, and had quite a number of visible holes, but the filtering and aggregation made up a bulk of the slack.
NetLens really stood out in my eyes, an 'Iterative Exploration of Content-Actor Network Data.' This presented the first really great interface of the day (aside from the previous presentation—Balancing Systematic and Flexible Exploration of Social Networks). The presentation was not overly crowded, and the interaction was rich. Really, the amount of possibilities this interface provided was stunning. Definately worth checking out.
