(Press right arrow or space bar to navigate to next slide)
What do we wish to achieve from network anonymization? Why do networks need to be anonymous? What are our constraints?
Netflix released an pseudonymized dataset of movie preferences, as part of a competition into recommender systems. Researchers use information from the Internet Movie Data Base (IMDB) to de-anonymize a number of records. After 2009 the completion is suspended following a lawsuit on privacy grounds.
De-anonymization attack components:
Should we be concerned? Is the method flawed? What would put our privacy at greater risk?