Twitter retweet predictor helps you track Kim Kardashian

A team of researchers at MIT has figured out how to predict how often a tweet will be retweeted. And face it: It’s all about the retweets. #lookhowcleverIam.

You can get an idea of what this all looks like here. (The blue line is the actual number of retweets. The red line is the predicted number.) If you want to go pedal-to-the-metal crazy on this, here is MIT Sloan School of Management Assistant Professor Tauhid Zaman etal’s paper.

It isn’t the first time MIT has polished its Twitter ball to peer into the future. Last year, the school created some buzz talking about buzz creation.

Now, before you accuse the very smart people at MIT of running out of smart things to study, the retweet-o-meter or the twouija (as in Ouija board, as MIT calls it) has all kinds of practical uses.

Some of these uses will be fun and could help you win bar bets, as MIT’s annoucement explains:

“‘Beyond a single tweet, understanding retweet behavior could lead to a better understanding of how broader ideas spread in Twitter and in other social networks,’ they write. ‘These ideas would consist of tweets from a large number of users on a similar topic. Understanding this type of information spreading would potentially allow one to predict which trends, memes, or ideas will become popular, how popular they will become, and how quickly they will become popular.”

Some of these uses seem destined to annoy us, like when marketeers get their hands on MIT’s model. Again from the announcement:

“Predictions of this sort have potential applications to marketing (new product adoption), politics (campaign effectiveness) and national security (protests and civil unrest), to name a few.”

Hoo-boy. I can hardly wait. Meatime, feel free to retweet this link liberally.

 

 

Mike Cassidy Mike Cassidy (173 Posts)

I write about the culture of Silicon Valley for the San Jose Mercury News.