Fingerprinting using database steganography

With the becoming of the age of data science, data as a primary resource requires protection, not only with respect to privacy. Still, many data driven technologies ask for the sharing of data with external experts, labs and other collaborators. Data fingerprinting and watermarking describes a set of techniques in order to protect the ownership of data even after sharing and enabling the identification of so-called data leaks, i.e. they make it possible to identify the leaking data recipient of a data set. In this work we provide a novel approach for watermarking whole databases based on steganographic methods that are robust against changes in the stored data set.