What happens when you crunch a heap of Facebook data with one of the most sophisticated computational tools in the world? Mathematical proof of the myriad ways your life gets boring as you get older. As you shed your youth, you can look forward to spending even more time whining about the weather and fretting about the government—and you'll make up the time by getting those pesky video games off your mind.
Mathematician Stephen Wolfram has started dissecting data for more than a million users using his sophisticated technology. Wolfram is probably best known for WolframAlpha, his "computational knowledge engine" which uses a sophisticated understanding of natural language to calculate answers to questions. Now he's trying to develop similarly robust "Personal Analytics".
In a blog post, Wolfram discusses the many layers of information he's been able to compute from the raw data. Much of it is a pretty theoretical analysis of how we connect with an emphasis on pretty standard demographic indicators like age and location.
It starts to get interesting when Wolfram starts to tackle more qualitative aspects of the data using natural language analysis, which he can use to figure out how your interests change over time. How does he do it? Once you've got a good engine for crunching language (as opposed to keywords) it's easy enough to figure out what broad topics people talk about the most. From there, Wolfram calculates with what frequency people in different age groups are talking about certain topics. He then plots those frequencies as trajectories over time. (In case it's not obvious, the red line refers to women and the blue line referes to men.)
And, man, isn't it just so hard not to read way too much into these numbers? Actually, not even. The charts are mostly surprising for how obvious the conclusions are. Nearly every chart passes the smell test, although, it's hard to try to explain it all without making some assumptions. Like it makes sense that we'd be more interested in inspirational quotes, philosophy, and in matters of health as we get older and are forced to confront our mortality. Is that fair?
These aren't perfect models for our interests, but with a little more crunching, you could easily see how that treasure trove we've been handing Facebook for years will eventually pay off. All along, you thought you were an individual, but it turns out that in some regards Facebook knows you better than you know yourself. [Stephen Wolfram via Technology Review]