Abstract: After analyzing the shortcomings of the existing text watermarking technologies,this paper presents a new incomplete semantically-based scheme.The watermark can be inserted into the content of the text without understanding the text completely.A feasible algorithm is given based on the auxiliary word “DE”.Furthermore,discussion on how to improve the capacity and the robustness of the algorithm are introduced.Experiments show that the embedding and the extraction of the watermark are easy to be realized and the algorithm itself is robust.

Key words: text digital watermarking, natural language processing, incomplete understanding of the text, DE phrase, watermark capacity