Linguistic Steganography Detection Algorithm Using Statistical Language Model

Steganography is a technique for embedding secret messages into carriers. Linguistic steganography is a branch of text steganography. Research on attacking methods against linguistic steganography plays an important role in Information Security (IS) area. In this paper, a linguistic steganography detecting algorithm using Statistical Language Model(SLM) is presented. An experiment to detect text segments generated by linguistec steganography systems NICETEXT, TEXTO and Markov-Chain-Based is carried out. The result of our experiment shows when the text segment size is 2K and 5K, the detecting accuracies are found to be 93.9% and 96.3% respectively