An Optimized Natural Language Watermarking Algorithm based on TMR

Peng Lu,Zhao Lu,Zili Zhou,Junzhong Gu
DOI: https://doi.org/10.1109/ICYCS.2008.398
2008-01-01
Abstract:Semantically-based watermarking scheme embeds the information by transforming semantic structure of the sentence. In this paper we propose an optimized semantic watermarking algorithm which introduces TMR string reranking mechanism before watermark embedding. The optimized algorithm has the feature of resisting sentence reordering attack to a certain extent. To evaluate the effectiveness of our algorithm, we conduct a few experiments with different conditions and make a comparison among the results. The experiments show that the optimized algorithm is effective and has better robustness.
What problem does this paper attempt to address?