Watermarking as a service (WaaS) with anonymity

Abstract

With rapid increase in Internet usage, it has become quite easy to produce illegal digital contents. As a result, copyright protection and authentication problems have risen. Besides audio, video and image, protection of database integrity is now a very active research in digital watermarking. Watermarking for databases has been introduced, however, most of the existing schemes produce distortion in the original content during watermark insertion phase. As a result of distortion to underlying data, the quality of watermarked content reduces. Moreover, most of the existing schemes provide watermarking techniques to protect specific data types, such as categorical or numeric. Furthermore, anonymity is also a critical challenge in prior watermarking techniques. In the said perspective, we propose watermarking as a service with anonymity for database integrity using zero watermarking schemes. Our proposed technique does not add any extra information, and neither any physical change is made to the database content. We also performed the security and performance analysis of our proposed scheme. For security analysis, we have evaluated against tuple insertion, tuple deletion, and attribute value tempering attacks. We can extract watermark with high probability, which shows resilience against the said attacks. Experiments are conducted to yield the satisfactory performance.