Better avoid any Freudian slips on Twitter!
TweetPsych, an app that uses two linguistic analysis algorithms to compare a person’s last 1,000 tweets against the ones in its database, now looks at lists.
The project uses the Regressive Imagery Dictionary, which measures primordial versus conceptual thinking, and the Linguistic Inquiry and Word Count, which looks for positive and negative emotions and references to topics like work, religion and eating. TweetPsych takes both of those approaches and can tell if your list tweets about “sex” more than the average one, for example.
Creator Dan Zarrella launched the service back during the summer, but he added lists today. As an example, I decided to psychoanalyze Venturebeat’s team list. In keeping with our brand, we are decidedly business-oriented. We tweet 187 percent more than the average list about ‘Media’, 113 percent more on ‘Money’, 44 percent more on ‘Numbers’ and 31 percent more on ‘Work’. We also tweet 50 percent less about sex. (Sorry, you’ll have to go somewhere else.)
Some food for thought: if there are fun, little free apps that do this already on the web, expect many more to crop up that do it for commercial purposes. Facebook, for one, already studies your stream data and status updates to help match advertisements.
We’ve written about projects that analyze tweets from the masses like FlixUp, which uses Twitter to gauge how good movies are, but we have yet to see any for-profit ventures that do deeper psychological profiling of users. That doesn’t mean it’s not out there though, so watch what you tweet!