Researchers have begun using Facebook as a social dataset for some very interesting research, including the recently released Gross National Happiness Index. The metric tracks aggregate "happiness" (based on the use of words like "happy," "joy," "awesome," etc) on a daily basis.
This is exciting news for me for two reasons. First, it means there are other people out there using these commercial websites to produce real research. It's validating to see others agree that online mega-sites are turning into social resources in their own right, spaces diverse and vast enough to support (more or less) general population research. Second, it's led me to LIWC, an intriguing piece of software for measuring different kinds of aggregate themes in texts--positive and negative emotions, for example. There are a number of similar efforts out there, but this one does seem to be fairly comprehensive, and it's been put to impressive use on the Facebook project. I'm thinking about how to analyze professional and consumer book reviews in more sophisticated ways and this route has some strong appeal.
