Personalization of Context Information Retrieval Model via User Search Behaviours for Ranking Document Relevance

Abstract

S techniques today are widely used for securing secret information. File types like image, audio, video and text are used for hiding secret information using steganographic techniques. A number of steganographic techniques are available in literature in which least significant bit (LSB) techniques are the most common. In this paper a new technique is demonstrated in which most significant bits (MSB) of pixels are taken into account for hiding secret data bit. In this technique secret data bits are arranged in pairs of two and similarly bits representing pixels are also arranged in pairs. For this purpose the 7th and 6th bits, 6th and 5th bits, 5th and 4th bits, 4th and 3rd bits are arranged to make pairs. The corresponding pairs are numbered as 3rd pair (7th and 6th), 2nd pair (6th and 5th), 1st pair (5th and 4th) and 0th pair (4th and 3rd). For each embedding the secret data bits pair is compared with the pairs of pixel bits and embedding is done on the basis of similarity between pairs. If no similarity is found in pairs then secret data pairs is embedded in 0th pair of pixel bits. The proposed technique ensures significant signal to noise ratio and is more secure than the existing techniques as the attackers mostly target the least significant bits of pixel for extracting secret information and focuses less on most significant bits.