A new website sponsored by Ubisoft
as part of its advertising campaign for the upcoming hacking-themed game Watch Dogs
isn't just a plug for the title -- it's a chilling example of exactly how easy it is for companies to mine your data. While I'm normally averse to giving any application or service access to my Facebook
account, I also take some effort to lock down my settings and posts. I decided to make an exception to this policy and give the application access to my account data. What it came back with is interesting, scary, and informative by turns.
There are precious few photos of me online, so I'm amused by Facebook's claim that it can identify me with 98.3% accuracy based on the following image:
That's my friend Ruth and her dog Rocky, a photo of a different dog, a black-and-white yearbook photo of me 18 years ago, and a photo of me as a kid holding some squirming cousin. Or maybe my sister. I don't recall and I refuse to take responsibility for that couch pattern -- it was the 1980s and a different time.
Datashadow also offers the ability to compare various character traits (shown here against Marco Chiappeta) and gives a great deal of information about total number of posts, post times (this appears to be factually accurate) and inferred values about income, location, and lifestyle.
It's just a game -- except it isn't
Is Ubisoft actually performing this kind of data analysis? Almost certainly not. This is far from an exhaustive, comprehensive examination of my personality or FB posting habits and the companies that actually perform that kind of data analysis are anything but cheap. The point
Ubisoft is making, however, is that your FB profile contains enormous amounts of information in a single place that can be mined in any number of ways.
Researchers have talked about this before, but boring sociological journals don't get the attention that a slick marketing campaign can generate. By analyzing your friends, your links, and the content of your posts, advertisers and FB itself can generate a profile of your political opinions and behaviors (or lack thereof). Even if you don't personally post much in the way of political material, your friends might -- and people with similar beliefs tend to cluster together.
The fundamental premise of much of Big Data
is that we can tease out particular trends and topics by looking at how people cluster together and self-identify. Do you posts stories about social, psychological, or political data? Do you follow religious groups on FB, or are you a fan of all of the various atheist organizations? Do you share links from science blogs or the likes of NaturalNews and anti-vaccination conspiracy theories?
All of this information absolutely is
combined and collated to create detailed digital profiles of all of us, and the more we engage with various online services (from Facebook to Google Plus), the larger the data pool becomes. There have been some discussions in the past around the idea that perhaps individuals deserve compensation for the way their data is fundamentally used and collated. While the Digital Shadow Value shown here is likely nothing but an in-game value, the idea that your own information is valuable-- and that you should be paid for it -- may gain ground in the future.