Natural Language Watermarking: Design, Analysis, and a Proof-of-Concept Implementation

By Mikhail J. Atallah, Victor Raskin, Michael Crogan, Christian F. Hempelmann, Florian Kerschbaum, Dina Mohamed, S. Naik. Information Hiding, 2001

308 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

LINGUISTIC STEGANOGRAPHY: SURVEY, ANALYSIS, AND ROBUSTNESS CONCERNS FOR HIDING INFORMATION IN TEXT (2004, methodology): “…with; Raskin and Nirenburg's system of ontological semantics has already been used to structure a watermarking system which takes advantage of the text structure of semantic representations in order to encode and restructure text in which bits are hidden (Atallah et al 2001, Atallah et al 2002).”
Provable Robust Watermarking for AI-Generated Text (2023, unknown): “Early approaches relied on techniques such as synonym substitution (Topkara et al., 2006), syntactic structure restructuring (Atallah et al., 2001), or paraphrasing (Atallah et al., 2002).”
Techniques and Applications of Digital Watermarking and Content Protection (2003, methodology): “While the previous method was developed for printed images, the following method was developed for binary digital images: Changing the ratio of pixels was proposed by Zhao and Koch [60].”
A Watermark for Large Language Models (2023, unknown): “3 , the average PPL of attacked sequences increases by 3 × in addition to requiring more model calls.”
Can AI-Generated Text be Reliably Detected? (2023, background): “Another line of work aims to watermark AI-generated texts to ease their detection [25, 26, 1, 27].”