31 October 2013
Facebook data scientists can figure out who you're sleeping with, even if your relationship isn't "Facebook official," and in a new paper researchers explain how they figured out the method, the Atlantic reports. Turns out the key is to look at more than just mutual friends. "Embeddedness," or the number of mutual friends two people have, can show how close those people are (the more mutual friends, the closer they are). But researchers found embeddedness predicted the correct significant others just 24.7% of the time. A more accurate predictor is something called "dispersion."
That means researchers looked at how many networks were shared between two people, working on the theory that your romantic partner will have met your family and your current friends and co-workers, as well as old friends from high school or college. By looking at which friend was the most dispersed across all of a Facebook user's networks, researchers upped their accuracy at predicting significant others to 50%. "A spouse or romantic partner is a bridge between a person’s different social worlds," one researcher explains, according to the New York Times. Oh, and dispersion also works to predict the health of your relationship: The more dispersion a user had with his or her partner, the more likely the relationship was to last at least 60 days.
Facebookis testing technology that would greatly expand the scope of data that it collects about its users, the head of the company’s analytics group said Tuesday.
The social network may start collecting data on minute user interactions with its content, such as how long a user’s cursor hovers over a certain part of its website, or whether a user’s newsfeed is visible at a given moment on the screen of his or her mobile phone, Facebook analytics chief Ken Rudin said Tuesday during an interview.
Mr. Rudin said the captured information could be added to a data analytics warehouse that is available for use throughout the company for an endless range of purposes–from product development to more precise targeting of advertising.
Facebook collects two kinds of data, demographic and behavioral. The demographic data—such as where a user lives or went to school—documents a user’s life beyond the network. The behavioral data—such as one’s circle of Facebook friends, or “likes”—is captured in real time on the network itself. The ongoing tests would greatly expand the behavioral data that is collected, according to Mr. Rudin. The tests are ongoing and part of a broader technology testing program, but Facebook should know within months whether it makes sense to incorporate the new data collection into the business, he said
New types of data Facebook may collect include “did your cursor hover over that ad … and was the newsfeed in a viewable area,” Mr. Rudin said. “It is a never-ending phase. I can’t promise that it will roll out. We probably will know in a couple of months,” said Mr. Rudin, a Silicon Valley veteran who arrived at Facebook in April 2012 from Zynga Inc., where he was vice president of analytics and platform technologies.(Zynga analytics is a data company masquerading as a gaming company)
Facebook also is a major user of Hadoop, an open-source framework that is used to store large amounts of data on clusters of inexpensive machines. Facebook designs its own hardware to store its massive data analytics warehouse, which has grown 4,000 times during the last four years to a current level of 300 petabytes. The company uses a modified version of Hadoop to manage its data, according to Mr. Rudin. There are additional software layers on top of Hadoop, which rank the value of data and make sure it is accessible.
Where does Facebook stop & the NSA begin:
Facebook (along with Google, Microsoft, etc.) was already collaborating with the National Security Agency's PRISM program that swept up personal data on vast numbers of internet users.
Facebook had to promise the feds it would stop doing things like putting your picture in ads targeted at your "friends"; that promise lasted only until this past summer, when it suddenly "clarified" its right to do with your (and your kids') photos whatever it sees fit. And just this week, Facebook analytics chief Ken Rudin told the Wall Street Journal that the company is experimenting with new ways to suck up your data, such as "how long a user’s cursor hovers over a certain part of its website, or whether a user’s newsfeed is visible at a given moment on the screen of his or her mobile phone."
It's your data that makes Facebook worth $100 billion and Google $300 billion. It's your data that info-mining companies like Acxiom and Datalogix package, repackage, sift, and sell. And it's your data that, as we've now learned, tech giants also pass along to the government.
Companies have given the NSA access to the records of every phone call made in the United States. Companies have inserted NSA-designed "back doors" in security software, giving the government (and, potentially, hackers—or other governments) access to everything from bank records to medical data. And oh, yeah, companies also flat-out sell your data to the NSA and other agencies.
Mark Zuckerberg said at this year's TechCrunch conference: The NSA really "blew it," he said, by insisting that its spying was mostly directed at foreigners. "Like, oh, wonderful, that's really going to inspire confidence in American internet companies. I thought that was really bad." Shorter: What matters is how quickly Facebook can achieve total world domination.
Maybe the biggest upside to l'affaire Snowden is that Americans are starting to wise up. "Advertisers" rank barely behind "hackers or criminals" on the list of entities that internet users say they don't want to be tracked by (followed by "people from your past"). A solid majority say it's very important to control access to their email, downloads, and location data. Perhaps that's why, outside the more sycophantic crevices of the tech press, the new iPhone's biometric capability was not greeted with the unadulterated exultation of the pre-PRISM era.
How much privacy we'll trade for either convenience or security—in someone else's hands: It's our responsibility to take charge of our online behavior (So long as you want your boss, and your high school nemesis, to see 'em), and, more urgently, it's our job to prod our elected representatives to take on the intelligence agencies and their private-sector pals.