TY - GEN
T1 - Effective linguistic steganography detection
AU - Zhi-li, Chen
AU - Liu-sheng, Huang
AU - Zhen-shan, Yu
AU - Xin-xin, Zhao
AU - Xue-ling, Zheng
PY - 2008
Y1 - 2008
N2 - Linguistic steganography is an art of concealing secret messages. More specifically, it takes advantage of the properties of natural language, such as the linguistic structure to hide messages. In this paper, an effective method for linguistic steganography detection is presented. In virtue of the concepts in area of information theory, the method uses an information-entropy-like statistical variable of words in detected text segment together with its variance as two classification features. The Support Vector Machine is used as classifier. The method was centered on detection for small size text segments estimated in the hundreds in words. Its achievement is simple and its execution is fast and relatively accurate. In our experiment of detecting the three different linguistic steganography methods: NICETEXT, TEXTO and Markov-Chain-Based, the accuracy exceeds 90%. As a result, our method can be used as a common pre-detection method followed by a more specific and accurate detection method.
AB - Linguistic steganography is an art of concealing secret messages. More specifically, it takes advantage of the properties of natural language, such as the linguistic structure to hide messages. In this paper, an effective method for linguistic steganography detection is presented. In virtue of the concepts in area of information theory, the method uses an information-entropy-like statistical variable of words in detected text segment together with its variance as two classification features. The Support Vector Machine is used as classifier. The method was centered on detection for small size text segments estimated in the hundreds in words. Its achievement is simple and its execution is fast and relatively accurate. In our experiment of detecting the three different linguistic steganography methods: NICETEXT, TEXTO and Markov-Chain-Based, the accuracy exceeds 90%. As a result, our method can be used as a common pre-detection method followed by a more specific and accurate detection method.
KW - Detection
KW - Entropy
KW - Information
KW - Linguistic steganography
UR - https://www.scopus.com/pages/publications/52049112347
U2 - 10.1109/CIT.2008.Workshops.69
DO - 10.1109/CIT.2008.Workshops.69
M3 - 会议稿件
AN - SCOPUS:52049112347
SN - 9780769533391
T3 - Proceedings - 8th IEEE International Conference on Computer and Information Technology Workshops, CIT Workshops 2008
SP - 224
EP - 229
BT - Proceedings - 8th IEEE International Conference on Computer and Information Technology Workshops, CIT Workshops 2008
T2 - 8th IEEE International Conference on Computer and Information Technology Workshops, CIT Workshops 2008
Y2 - 8 July 2008 through 11 July 2008
ER -