Anomaly Data Leakage Detection
B. Chellaprabha · 2013
In course of doing business the data should be shared between the other organizations. The owner of the data are said to be Distributor and the trusted agents are said to be Agents. The Distributor will give some of the sensitive data to the agents which are created by them for the manipulation process. The agents are created with individual id and they are allocated with the data process. While processing the data there may be a chance of leaking those data to other unwanted agents or persons. Here, data leakage can be detected or prevented by using the new Online learning ensemble approach called Concept Drift. Hence the system is designed in such a way that the data on reaching any destination say agent or unauthorized party, data as well as the IP address of the receiver will reach the distributor. If the distributor receives an IP address other an agent's address, the distributor finds out that the data has been leaked. Fake object is inserted along with the data at the time of distribution for detecting the guilty agent. From the data received, the distributor compares with his database and finds out which agent has leaked the data by calculating the probability for each agent on the leaked data.