There is no such thing as complete anonymity

Unmasking Social-Network Users

University of Texas at Austin researchers have found that, combined with readily available data from other online sources, social network data can reveal sensitive information about users. Using the photo-sharing site Flickr and the microblogging service Twitter, the researchers were able to identify a third of the users with accounts on both sites by searching for recognizable patterns in anonymized network data. Both Twitter and Flickr display user information publicly, so the researchers anonymized much of the data to test their algorithms.

The objective was to determine if it was possible to extract sensitive information on individuals using the connections between users, even if almost all of the personally identifying information had been removed. The researchers found that extracting information was possible provided they could compare patterns with those from another social-network graph in which some user information was accessible.

Texas professor Vitaly Shmatikov notes that social network data, particularly the patterns of friendships between users, can be valuable to advertisers. However, he says releasing that information also makes the networks vulnerable. The researchers found that non-anonymous social network data is easy to find. Every person does a few quirky, individual things which end up being strongly identifying. Carnegie Mellon University professor Alessandro Acquisti says the research points to the difficulty in maintaining privacy online. "There is no such thing as complete anonymity," Acquisti says. "It's impossible."

Refer here to read more details about this research.