Attacks on Translation Based Steganography

Author

Meng, Peng ; Hang, Liusheng ; Yang, Wei ; Chen, Zhili

Author_Institution

Depart. of CS.&Tech., USTC, Hefei, China

fYear

2009

fDate

20-21 Sept. 2009

Firstpage

227

Lastpage

230

Abstract

Translation-based steganography (TBS) is a kind of famous text steganography. In this paper, we examine the robustness of TBS and give an effective detection algorithm for TBS. Our algorithm can not only distinguish between natural language text and stego-text which was generated by TBS, but also can distinguish between machine translated text and stego-text. The detection accuracy increases as the text size increases. When the text size is larger than 60 sentences, the detection accuracy is greater than 92%.

Keywords

language translation; natural language processing; steganography; text analysis; detection algorithm; machine translated text; natural language text; stego-text; text steganography; translation based steganography; Cryptography; Detection algorithms; IEEE members; Image reconstruction; Natural languages; Protocols; Robustness; Steganography; Support vector machine classification; Support vector machines; DMP(Degree of Machine Preference); DMR(Degree of Machine Reversibility); Steganalysis; TBS(Translation-Based Steganography); Text Steganography;

fLanguage

English

Publisher

ieee

Conference_Titel

Information, Computing and Telecommunication, 2009. YC-ICT '09. IEEE Youth Conference on

Conference_Location

Beijing

Print_ISBN

978-1-4244-5074-9

Electronic_ISBN

978-1-4244-5076-3

Type

conf

DOI

10.1109/YCICT.2009.5382382

Filename

5382382