Research on Coverless Text Steganography Based on Multi-Rule Language Models Alternation

Information hiding technology usually uses video, audio, image, and text as carriers. Among them, text redundancy is less and it is difficult to hide information. However, the text is widely used, so text-based information hiding has very important research significance. In this paper, a coverless text steganography based on multi-rule language models alternation is proposed. It is different from the traditional steganographic text generation method based on a single gram model. The method proposed in the paper tries to combine models under different language rules alternately, and tries to extract more language features from the training text. In order to verify the effectiveness of the method, we design some relevant experiments. The results show that, compared with other text steganography methods, the method proposed in the paper not only satisfies the concealment but also improves the hiding capacity.