Linguistic steganography: state-of-the-art. Part one

In the article the notion of linguistic steganography is positioned amongst the system of information hiding methods as a means to generate a (pseudo-) text with the embedded covert message. The notion of linguistic steganography proper, implying mimicking of linguistic features of a “natural” human-made text when composing a stegotext, and text steganography, including format-based transformations of the cover text, are opposed. The levels of steganographic text modeling are distinguished. These levels are accounted for when generating a stegotext characterized by the feature distribution typical of a non-modified text in a natural language. The techniques used to model statistical and syntactic features of the text are considered. Advantages and disadvantages of the techniques presented are analyzed.