Research by Carnegie Mellon professor Latanya Sweeney and other experts shows that an alarming number of seemingly innocuous, neutral, or "common" data points, can potentially identify an individual online.