Natural Language Watermarking and Tamperproofing

Atallah, Mikhail J.; Raskin, Victor; Hempelmann, Christian F.; Karahan, Mercan; Sion, Radu; Topkara, Umut; Triezenberg, Katrina E.

doi:10.1007/3-540-36415-3_13

Cited by 78 publications

(80 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Natural language watermarking embeds information using linguistic transformations such as synonym substitution, syntactic transformations and semantic transformations (see [18], [17], [2], [19], [1], [8], [10]). It embeds a bitstream as a watermark in a text document, while preserving the meaning, context, and flow of the document.…”

Section: Natural Language Watermarkingmentioning

confidence: 99%

“…There have only been a few such schemes, e.g., [2], [1] and [8], that are suitable for this purpose. In [2] Atallah et al treat a text as a collection of sentences.…”

Section: Previous Schemesmentioning

confidence: 99%

“…A follow-up paper [1] improves upon [2] in terms of the number of watermark bits embedded in one sentence. The algorithms used in [1] are the same as in [2], but the algorithms are applied to the sematic structure of a sentence, instead of the syntactic structure, to embed the watermark bits.…”

Section: Previous Schemesmentioning

confidence: 99%

“…The algorithms used in [1] are the same as in [2], but the algorithms are applied to the sematic structure of a sentence, instead of the syntactic structure, to embed the watermark bits. Since the semantic structure is much larger and richer than the syntactic structure, this allows transformations of a large number of elements they contain, and therefore it achieves more watermark bits in one sentence.…”

Section: Previous Schemesmentioning

confidence: 99%

“…Since the semantic structure is much larger and richer than the syntactic structure, this allows transformations of a large number of elements they contain, and therefore it achieves more watermark bits in one sentence. In view of robustness, the probability that one sentence insert/delete/modify can destroy the watermark in [1] is the same as [2], therefore its false negative probability is also the same as [2].…”

Section: Previous Schemesmentioning

confidence: 99%

See 4 more Smart Citations

Authorship Proof for Textual Document

Stinson

2008

Information Hiding

View full text Add to dashboard Cite

Abstract. In this paper, we investigate the problem of how to prove the authorship of textual documents. We propose to use natural language watermarks to solve the problem. We identify two essential requirements for an authorship proof scheme (APS) to be secure against watermark erasing attacks. We evaluate the security of existing natural language watermarking schemes, and we propose two new APS with improved security.

show abstract

Section: Natural Language Watermarkingmentioning

confidence: 99%

“…There have only been a few such schemes, e.g., [2], [1] and [8], that are suitable for this purpose. In [2] Atallah et al treat a text as a collection of sentences.…”

Section: Previous Schemesmentioning

confidence: 99%

Section: Previous Schemesmentioning

confidence: 99%

Section: Previous Schemesmentioning

confidence: 99%

Section: Previous Schemesmentioning

confidence: 99%

See 3 more Smart Citations

Authorship Proof for Textual Document

Stinson

2008

Information Hiding

View full text Add to dashboard Cite

show abstract

Matlist: Mature linguistic steganography methodology

Desoky

2011

Security Comm Networks

View full text Add to dashboard Cite

The generated text by Natural Language Generation (NLG) and template systems is meaningful and looks legitimate. Therefore, the Mature Linguistic Steganography Methodology (Matlist) employs NLG and template techniques along with Random Series values (RS), e.g. binary, decimal, hexadecimal, octal, alphabetic, alphanumeric, etc., of Domain-Specific Subject (DSS) to generate noiseless text-cover. This type of DSS, e.g. financial, medical, mathematical, scientific, economical, etc., has plenty of room to conceal data and allows communicating parties to establish a covert channel such as a relationship based on the profession of the communication parties to transmit a text-cover. Matlist embeds data in a form of RS values, function of RS, related semantics of RS, a combination of these, etc. Unlike synonym-based approach, Matlist does not preserve the meaning of text-cover every time it is used. Instead, Matlist Cover retains different legitimate meaning for each message while it remains semantically coherent and rhetorically sound. The presented implementation, validation, and experimental results demonstrate that Matlist is capable of accomplishing the steganographical goal with higher bitrate than all other linguistic steganography approaches.

show abstract

Adaptive‐capacity and robust natural language watermarking for agglutinative languages

Kim

Goebel

2011

Security Comm Networks

View full text Add to dashboard Cite

We present a robust and adaptive‐capacity watermarking algorithm for agglutinative languages. All processes, including the selection of sentences to be watermarked, watermark embedding, and watermark extraction, are based on syntactic dependency trees. We show that it is more robust to use syntactic dependency trees than the surface forms of sentences in text watermarking. For the agglutinative languages, we embed watermark using the two main characteristics of the languages. First, because a word consists of several morphemes, we can watermark sentences using morphological division/combination without deep linguistic analysis. Second, they permit relatively free word order, so we can move a syntactic constituent within its clause. Finally, to increase the information‐hiding capacity, we adaptively compute the number of watermark bits to be embedded for each sentence. We perform three kinds of evaluation: perceptibility, robustness, and capacity of our method. High capacity is achieved by dynamically determining possibly embedded watermark bits for each sentence. The secret rank based on a syntactic dependency tree strengthens robustness of our method. Finally, we show that the displacement of syntactic constituents and morphological division/combination does not affect the style and naturalness of the text. Copyright © 2011 John Wiley & Sons, Ltd.

show abstract

Natural Language Watermarking and Tamperproofing

Cited by 78 publications

References 7 publications

Authorship Proof for Textual Document

Authorship Proof for Textual Document

Matlist: Mature linguistic steganography methodology

Adaptive‐capacity and robust natural language watermarking for agglutinative languages

Contact Info

Product

Resources

About