Many of the analytic applications built on twitter measure authority, reach, followership, etc. They basically examine the social graphs of users. It seems to me that not enough is being done to address the temporal aspects of twitter. If twitter is truly the real-time web, and the value of the stream is largely in its dynamics, I’m suprised there aren’t more applications out there focusing on twitter’s temporal dynamics — not just how many re-tweets, but when do those re-tweets happen? How are they grouped in time? What’s the inter-tweet time? Who are the early re-tweeters (twitter’s first responders…)? How rapidly do ideas diffuse through different parts of the graph? What time-of-day do ideas spread the fastest?
I’d love to get a better sense for the impusle response of twitter. Each tweet is a probe of a dynamic system and the temporal characteristics of the after-effects contain a great deal of information. It’s different for each user, and I suspect that the ones with the most responsive ‘channels’ are not necessarily the ones with the most followers. Maybe I’ll build a system to capture each user’s impulse response. Maybe somebody already has?