Wolfram Alpha Drills Deep into Facebook Data

Grapes? No, a visualization of peoples’ Facebook “friends” networks.

Back in January, when Wolfram Alpha launched an updated version of its Personal Analytics for Facebook module, the self-billed “computational knowledge engine” asked users to contribute their detailed Facebook data for research purposes.

The researchers at Wolfram Alpha, having crunched all that information, are now offering some data on how users interact with Facebook. For starters, the median number of “friends” is 342, with the average number of friends peaking for those in their late teens before declining at a steady rate. Younger people also have a tendency to largely add Facebook friends around their own age—for example, someone who’s 20 might have lots of friends in the twenty-something range, and comparatively few in other decades of life—while middle-aged people tend to have friends across the age spectrum.

“The first thing we see is that the ages of friends always peak at or near the age of the person themselves—which is presumably a reflection of the fact that in today’s society many friends are made in age-based classes in school or college,” read an April 24 note on the Wolfram Alpha blog. “For younger people, the peak around the person’s age tends to be pretty sharp. For older people, the distribution gets progressively broader.”

Facebook relationship status also shifts as people age. The number of those self-reporting as “single” starts off as a healthy percentage in the teens and twenties before plunging to the minority in the thirties and forties, replaced by those reporting as “married,” “engaged,” or “in a relationship.” In an interesting twist, the number of those reporting “Other” (whatever that means) stays fairly constant throughout the sample’s collective age range, while there’s a spike in those reporting as “widowed” near the older end:

“The fraction of people who report themselves as married continues to increase roughly linearly with age,” the blog noted, “gaining about 5 percent between age 40 and age 60—while the fraction of people who report themselves as single continues to increase for women, while decreasing for men.”

Researchers found that Facebook’s curve for married vs. age mirrored that of the U.S. Census:

Meanwhile, Facebook’s population was generally younger than that of the actual United States:

Beyond that, the Wolfram Alpha blog offers up some interesting information about friend counts (and “friend of friend” counts), how friends’ networks tend to “cluster” around life events such as school and sports teams, and even how peoples’ postings tend to evolve as they get older—as people age, for example, they tend to talk less about video games and more about politics.

“It feels like we’re starting to be able to train a serious ‘computational telescope’ on the ‘social universe,’” the blog concluded. “And it’s letting us discover all sorts of phenomena.” Of course, if Wolfram Alpha can break down data to this degree, it stands to reason that advertisers and corporate researchers can, as well—hinting at how those entities might try to slice and dice data in order to do everything from sell products to determine insurance rates.

Launched in 2009, Wolfram Alpha is designed to give users definitive (and usually numerical) answers in response to queries, making it an altogether different type of search engine from Google or Bing. Wolfram Alpha’s creators first launched its Facebook data-mining module in August 2012, allowing the social-network’s users to leverage the platform’s computational abilities to analyze and visualize their weekly distribution of Facebook posts, types of posts (photos, links, status updates), weekly app activity, frequency of particular words in posts, and more.

At this year’s South by Southwest (SXSW) conference in Austin, Texas, Wolfram Alpha creator Stephen Wolfram offered up some interesting details about his computational engine. Wolfram Alpha contains more than 10 trillion pieces of data cultivated from primary sources, along with tens of thousands of algorithms and equations. Solving complex math problems is one of the system’s key abilities. In the future, Wolfram anticipated that Wolfram Alpha would become more anticipatory of peoples’ queries, but he didn’t elaborate overmuch on that particular point.


Images: Wolfram Alpha