A Novel Approach for Preserving Privacy of Content Based Information Reterival System

Abstract

Content based information retrieval system (CBIR) are advanced version of retrieval systems where search is based upon specific criteria in order to get relevant items. In networking environment, as search is based on content it is easy for server to know client’s interest, where client has to trust server to get relevant items. Sometimes query contains sensitive information that client does not want to reveal it, but still search should be performed. This is achieved by our proposed structure, where mainly it will deal with multimedia items such as image or audio files. In order to preserve privacy , client selects multimedia file of which hash value is generated, this value is fired towards cloud server. Cloud server contains database of stored hash values of multimedia items and based upon hamming distance and similarity search, encrypted candidate list is prepared and send it to client. Client finds best item by carrying decryption.