Detecting Data Leaks
This project addresses various data leakage problems: Consider a data sender who transfers some confidential data to some of his counterparts (third party). Some of this sensitive data gets leaked accidentally or purposely by an attacker and is downloaded on his terminal. The distributor must find the possibilities that the compromised data was from one or more of his counterparts, as opposed to having gathered by other means. So our project allows for data allocation tactics (through the counterparts) that improves the chances of identifying data leakages. These methods do not depend on various alterations of the transferred data like parity. In some of the cases, we can also insert ?realistic but fake? data records into the data stream to further improve our chances of identifying unknown data leakages and also the party responsible for it.