The software complex for attack the linguistic stegosystem

Abstract

The article reviews the common methods and algorithms of textual steganalysis and text summarization, as well as automated computer systems based on them. It is described the disadvantages for accomplishing the task of the linguistic stegosystem attacking by semantic compression in order to remove a hidden message. The article describes the development and pre-testing of the software complex, based on the semantic compression of the text for attacking the linguistic stegosystem and taking into account the initial structure of this text. The development is aimed at detecting and neutralizing the threats caused by the use of computer linguistic steganography methods. Described program complex consists of 9 interconnected modules, 25 systems and 6 subsystems. Their interaction is highlighted and the system’s algorithm is shown in the article. The pre-testing results on the basis of the material consisting of 1000 generated texts, divided into such two categories, as the stegomessage hiding method and the volume of text, prove the effectiveness of the attack using the proposed program. The stegomessage may be deleted by 97%, the average compression ratio is 2.5 for single sentences and 4.6 for solid texts. During the preliminary testing it was found the main advantages of the system, which include the automation the process of attacking a text-based stegosystem using both meaningful and meaningless texts and a wide range of research in stegoanalysis. Thus, the efficiency and practical capacity of the developed author's method, aimed at the textual data semantic compression for the purpose of counteracting the methods of linguistic steganography, has been proved. The further studies will be related to elimination the disadvantages detected during the preliminary testing and to confirming the preliminary conclusions of the work.