A robust and reversible watermarking technique for relational dataset based on clustering

With rapid information development, data sharing becomes a crucial part in the Internet. In the process of sharing, data ownership protection and data traceability are two key issues that need to be solved urgently. To address these problem, digital watermarking technology can be a solution. Digital watermarking is used to guard the rights of owners of digital products. Many robust and reversible watermarking techniques are proposed recently to ensure the rights and recover original data set. But most methods require primary keys of the data as a required parameter, resulting in original data not recovered and partial data not traceable against data structure attack. In this paper, a cluster-based robust and reversible watermarking (RRWC) technique for relational data has been proposed that provides a solution to two major function: ownership rights protection and partial data traceability. The unsupervised classification algorithm is used to group dataset, where the primary key of the data will not be used and the watermarks can be embedded with low distortion and high capacity. RRWC addresses malicious attacks, such as subset insertion attack, deletion attack, alteration attack and data structure attack. Experimental results demonstrate the effectiveness and robustness of RRWC against attacks.