You’ve gotta give Facebook credit for its relentless drive to harvest and exploit the data of its billion users. Facebook knows where you are, where you’re from, the places you Like, the people you know, and what you’re interested in — as soon as it makes its way to wearable computers it’ll probably know when you’re sleeping or awake, like a technologically twisted version of Santa Claus. The company is making some of that data easier to discover with Graph Search, a social search tool that will begin rolling out to all US-based Facebook users today.
Graph Search is advertised as a tool that will make it easier to discover your friends’ interests, the places they check-in to, their photos, and “friends of friends.” It’s also the closest many people will come to learning just how much personal information they’re sharing without even realizing it. Given the recent reports concerning PRISM, an initiative through which the National Security Agency is able to collect data (or metadata) from companies like Facebook, Apple, Microsoft, and others, this demonstration of Facebook’s knowledge should offer the masses some idea of the extent to which their privacy has been invaded.
Put another way: Facebook has built a consumer-focused tool that highlights just how much information can be gleaned from seemingly worthless data. And it’s doing so just weeks after the Guardian reported that it was helping the US government spy on, well, everyone. That takes some gall.
The current iteration of Graph Search relies on data knowingly provided by Facebook’s users. The company can’t yet divine your birthday, hometown, or work on its own; you have to tell it all of those things, and, if you’re feeling charitable, what kind of music you like, what movies you’ve seen recently, who you’re dating, and where you are at any given moment. This might change in the future, as the New York Times reports that Facebook is working to improve Graph Search’s results by parsing the contents of individual Status Updates and data from third-party services like Yelp or Instagram. You already tell Facebook so much — you’re about to tell it so much more.
And that’s just the data you know you’re sharing. Every time you use Facebook, however, you’re also generating a whole other class of data called metadata — literally “data about data” — that is often shared without your consent or knowledge. You automatically share your name, your birthday, your location, and the device you’re using to access Facebook, according to the Guardian; some of these, such as your location, can be blocked on the device level, but others are attached to everything you share whether you like it — or know it — or not.
Immersion, a tool developed by a team within MIT Media Lab, shows how metadata can be used to glean information about a subject. In Immersion’s case the data is sourced from Gmail (you can check the data associated with your account at the tool’s website) and simply shows who you’ve been emailing, how those people are connected to each other, and how often you interact with each. Even such a limited tool is able to show the connections between Gmail users — and with metadata, as with life, so much of it is simply about who you know.
Much more could be gleaned from Facebook’s metadata. Someone might be able to find out how often you go to a particular movie theater, even if you never check-in. Who you’re with every night, even if neither of you tags the other in a Status Update. Where you’re from, where you’ve been, and with whom you’ve travelled. All of this data is currently stored on Facebook’s servers as a mind-bogglingly large number of disparate dots; imagine what might be gleaned if someone were able to connect even a fraction of them.
Graph Search shows what can be done when the dots associated with publicly-available information are connected. It should also serve as an example of what someone with the proper motivation could do if they were to access all of those public dots and connect them to other data that you probably didn’t even know you were sharing. Facebook built Graph Search to prove that Likes could be worth something, to offer yet another marketing tool to its advertisers, and to show that it was more than a time-suck devoted to “Candy Crush” and pictures of an acquaintance’s Spring Break. It also accidentally offered the average person a look at just how much they’ve shared with the service over the years.
[Illustration by Hallie Bateman]