Combined Markov Model and Zero Watermarking for Integrity Verification of English Text Documents

Abstract

To assure the integrity of the content by detecting attempted tampering of the original document is integrity verification attempt. The documents are usually watermarked with a watermarking technique. The tampering in document can alter or destruct the watermark and can misuse the document by the unauthorized user. Zero watermarking provides the most accurate results for tampering detection and integrity verification. The watermarking technique has some stages including watermark embedding, watermark extraction, watermark detection, and possible attacks during the watermarking process. The watermark embedding process performed by the embedding algorithm and the watermark extraction process extracts the watermark from the text by extraction algorithm. Combined Markov model and zero watermarking for integrity verification of English text documents are presented in this paper. The letter level- and word level-based combined Markov model and zero watermarking method are analyzed. 3 g LNMZW and 5 g WNMZW algorithms are developed. Based on the N-gram orders of Markov model in word mechanism, each word of the text is considered as 1 g or one order of Markov model based on this procedure the 5 g of Markov model is improved and compared its performance with other algorithms with different attacks in this paper.