English

Frustratingly Easy Edit-based Linguistic Steganography with a Masked Language Model

Computation and Language 2021-04-21 v1

Abstract

With advances in neural language models, the focus of linguistic steganography has shifted from edit-based approaches to generation-based ones. While the latter's payload capacity is impressive, generating genuine-looking texts remains challenging. In this paper, we revisit edit-based linguistic steganography, with the idea that a masked language model offers an off-the-shelf solution. The proposed method eliminates painstaking rule construction and has a high payload capacity for an edit-based model. It is also shown to be more secure against automatic detection than a generation-based method while offering better control of the security/payload capacity trade-off.

Keywords

Cite

@article{arxiv.2104.09833,
  title  = {Frustratingly Easy Edit-based Linguistic Steganography with a Masked Language Model},
  author = {Honai Ueoka and Yugo Murawaki and Sadao Kurohashi},
  journal= {arXiv preprint arXiv:2104.09833},
  year   = {2021}
}

Comments

7 pages, 4 firgures

R2 v1 2026-06-24T01:21:39.824Z