Natural language watermark robustness testing method and system

Abstract

The invention discloses a natural language watermark robustness testing method which comprises the following steps of: reading parameters required for system operation; based on a watermark embedding algorithm to be tested, embedding watermark information into an original carrier text to generate a watermark text; based on active attack methods, performing an active attack on the watermark text to generate an attacked text; based on a watermark extraction algorithm to be tested, extracting new watermark information from the attacked text; and calculating an error bit rate. According to the invention, different active attack methods are obtained according to an adversarial model of the watermark algorithm, the watermark text is attacked based on the active attack methods, the new watermark information is then extracted from the attacked watermark text, and the error bit rate is finally obtained. The experiment proves that the method disclosed by the invention can effectively and objectively evaluate the robustness of a natural language watermark algorithm to be tested.

Description

technical field [0001] The invention relates to the field of natural language watermarking, in particular to a natural language watermarking robustness testing method and system thereof. Background technique [0002] With the development of computer technology and the popularization of the Internet, on the one hand, people enjoy the convenience brought by the development of science and technology, but on the other hand, they are more and more worried about the safe transmission of information. Whether it is national military information, literary works or personal privacy, people hope that these information can be transmitted safely on the network, so the technology of digital watermarking has attracted more and more attention and attention. In 1993, Tirkel officially proposed the concept of digital watermark for the first time. Digital watermark aims to protect the copyright of digital works by embedding watermarks in images, audio and video works. [0003] Image watermark...

AI Technical Summary

Problems solved by technology

However, the use of image watermarks to protect text works faces the following problems: (1) Works must be saved and disseminated in the form of pictures, which prevents end users from editing normally; (2) OCR can remove watermarks, resulting in invalid protection; (3) The dissemination form of the picture is likely to arouse the suspicion of the opponent, which increases the possibility of the watermark being destroyed

[0005] At present, there is no unified evaluation scheme for the robustness of natural language watermarking, so various watermarking algorithms cannot be compared fairly, which leads to: (1) users cannot know whether their basic needs have been met; (2) researchers Unable to propose improvements to one's own algorithm by comparing different algorithms

In addition, researchers are not necessarily comprehensive and objective in evaluating their proposed watermarking algorithms, and many algorithms that claim to be safe have already been broken.

[0006] To sum up, although natural language watermarking can protect the copyright of written works, there is currently a lack of a unified evaluation scheme for the robustness of natural language watermarking, which restricts the application and scientific evaluation of natural language watermarking technology

Images