Data-mining 10,000 Facebook status updates to graph breakup patterns
geekosystem.comIt's been said that the credit card companies have such good data on spending patterns that they can predict major life events, such as weddings and divorces.
Anyway, I called Amex and they wouldn't tell me mine...
original post: http://mathiasmikkelsen.com/2010/10/amazing-facts-about-face...
although still no y-axis :(
I'm curious to see if this guy will get a letter from Facebook's legal department, like Pete Warden did.
Based on their election widget and access to demographic information, Facebook probably could put together scarily accurate, realtime election results.
This is quite US-centric.
The world is US-centric.
Facebook isn’t. From http://www.facebook.com/press/info.php?statistics : “About 70% of Facebook users are outside the United States”
But it isn't the US vs the World, its the US vs other individual countries. Which other countries make up 30% of Facebook's population?
This means 10000 status updates that mentioned a break-up. I did not guess that at first and was surprised how many of 10000 random updates could have been breakups...
Probably looked at relationship status changes. I'd personally be interested in seeing the opposite -- i.e., peaks in people starting relationships, as well as the inter-arrival periods of relationships.
Hm. Any idea why people break up so often before Christmas?
"I'm thinking about breaking up anyway, and I don't feel like spending a bunch of money on a Christmas gift."
Just a guess: because it reminds you that another year just passed ?
If you don't like your current situation, then it put you into the perspective of your whole life being wasted.
College kids rushing into relationships at the start of term and then breaking up at Christmas when they go home for holidays?
10,000 statuses over the course of year is not a significant percentage. Did he just pool his friends status's?