Finding needles in the social media haystack.

It's a cold, blustery day outside so I'm huddled in my office with a hot cup of coffee catching up on a week's worth of reading. One article at the top of my list this week is  this short piece on Trends and Outliers, the blog of TIBCO's Spotfire data visualization software. In a nutshell, the post outlines efforts by researches to use Twitter information as the basis for predictive models. Professors at MIT have created a model that they say can predict hot topics before they go viral, while a researcher at UC Riverside is building a model that forecasts stock prices based on Twitter chatter about various firms. 

Interesting stuff, with potentially fantastic implications for fundraising. Imagine being able to shift through a pile of tweets to find donors more likely to give at year end. And at the same time, as seems to be the case with all applications of predictive modeling, I see sinister undertones as well. Do we want our global economic health, for example, to be dictated by the whims of millions of Twitter users? Although I guess one could argue that we're not far from that reality already...

In any case, worth a few minutes of your time. I hope you are warm and cozy wherever you are!