Document Database Watermark Algorithm Based on Connected Graph

Abstract

With the wide application of Web technology and big data, non-relational database (NoSQL database) occupies an increasingly important position. Document database is the most widely used NoSQL database. In order to solve the problems such as copyright protection, data authentication and so on, a document database watermark scheme based on the connected graph is proposed. According to the real documents in the database, the pseudo documents are generated by genetic algorithm, and the watermark is embedded in the modifiable fields of the pseudo documents. The fitness function ensures the integration of real and pseudo documents. So that attackers cannot distinguish between real and pseudo documents, which increases the difficulty of attack. By forming pseudo documents into k-connected graphs with pointer fields, authorized users can distinguish between real and pseudo documents through connectivity. Experimental results show that the scheme achieves good invisibility and robustness.