Synonym substitution-based steganographic algorithm with vector distance of two-gram dependency collocations

In Linguistic steganography, when secret information is embedded into to the text using synonym substitution-base method, obvious mistakes and logical misconceptions, resulted from the inaccuracy of candidate synonyms, are ubiquitous. A new steganography algorithm is proposed based on vector distance of two-gram dependency collocations. The algorithm can calculate the appropriateness of synonym substitution. Firstly, the synonym which has the same attribute and semantic as the target word is chosen from the WordNet sets. Secondly, the two-gram dependency collocations is should be extracted from the target sentence using dependency syntactic, and the best synonym substitution set is obtained by constructing and calculating vector distance of two-gram dependency collocations in a large-scale corpus. Experimental results show that the stego text generated using the algorithm keeps the property of the text the same. In comparison to the current advanced synonym substitution, not only can it better ensure the accuracy of grammar and completeness of semantics, but it is able to resist attacks made by detection algorithms using statistical features based on the synonyms pairing and relative frequency more efficiently, guaranteeing the security of secret information.