A new study examines how organizations use information people disclose on social network sites (SNS) to predict their personal characteristics and whether SNS users can successfully block certain information (and how much) to better protect their privacy. A novel analytical tool called a "cloaking device" to prevent the use of specific information and how effective it may be are discussed in an article in Big Data.
The article entitled "Enhancing Transparency and Control When Drawing Data-Driven Inferences About Individuals" is coauthored by Daizhuo Chen, Columbia Business School (New York, NY), Samuel Fraiberger, Northeastern University (Boston, MA), and Robert Moakler and Foster Provost, Stern School of Business, New York University. They focused on the types of inferences about individuals that can be made based on their "Likes" on Facebook. They describe how organizations can be more transparent about how they use information from SNS to make personal inferences. The researchers introduce the "cloaking device" they developed and discuss how much information users need to cloak to have a significant effect on its predictive value.
"This is a landmark article," says Big Data Editor-in-Chief Vasant Dhar, Professor at the Stern School of Business and the Center for Data Science at New York University. "Given how routinely social media sites violate individual privacy for targeting, it is important for end users to get back some control over the kinds of things that are inferred about them from their surfing behavior. This paper provides a practical model for how users can cloak their identity and avoid certain types of inferences to be drawn about them."
Explore further: When is big data too big? Making data-based models comprehensible