A Survey Paper on Data Leak Detection using Semi Honest Provider Framework

Current Statistics from various security organization research firms and government institutes suggest that there has been a rapid growth of data leak in past 8 years. Thera are various reasons for data leaks amongst which human errors is most endorsing of all. There are many ways in which we can have an audit trail verifying for data leak in networks, but still not all consider human errors as an important check factor, which makes them more prone to fail. Some of the common solutions include keeping a copy of sensitive data at the providers end and maintaining an audit trial for checking whether there is any data leakage in networks, which in turn notifies the organization about data leakage. All these methods in turn are prone to data attacks, as the keep copy of data while auditing. So keeping these errors in mind, we present a novel data leak detection (DLD) framework where we maintain a set of important data or sensitive data digests which is used for the purpose of detection of any data leakage in the network. So this method proves to be non-erroneous due to the fact that the owner does not completely have to expose the sensitive data to the semi honest provider for data leak detection method. The owner has data control rights i. e. he can decide what part of data to be revealed to the semi honest provider and which not to reveal.