Research On Text Data Integrity Detection Scheme In The Outsourced Database Model

Database service outsourcing is a new database application paradigm that has been proposed recently. In the outsourced database model, the data integrity detection is to prevent the databases tampered with from malicious database service providers, and to assure queriers that the query results have not been tempered with and are authentic with respect to the actual data owner. Previous work on database outsourcing has focused on the structured data in the relations database, while as for unstructured data there is not a good solution at present. For perfecting the data integrity detection technology in the outsourced database model, in this work a text data integrity detection scheme based on the digital fingerprint is presented.Firstly, the basic working principle of the outsourced database model is introduced. Then, the Merit and Shortcoming of the existing data integrity detection technology as well as the Special safety mechanism of the outsourcing database are analyzed. Furthermore, the relative technologies and the major problem about Chinese word segmentation are analyzed.Secondly, in order to increase the speed of Chinese word segmentation, according to Chinese word characteristics, a new dictionary is established and a Chinese word segmentation method through divide-and-conquer strategy is presented, and the overlapping ambiguity in Chinese words segmentation is solved using statistical techniques.Thirdly, through analyzing the text data, the extraction method of digital fingerprint is proposed by using the Karp-Rabin algorithm thought. Based on above a fingerprint-based integrity detection scheme for text data is presented, this scheme not only detected out the text data integrity, but also the data which had been tampered with can be located exactly. Moreover, the client and server fingerprint management plans, whose security is analyzed, are given. The integrity detection scheme of the cryptographic text data and the plain text data in the above situation is discussed respectively.Finally, the word segmentation method and text-data integrity algorithm is analyzed theoretically. Furthermore, the experiments verification of the correctness and effectiveness are done.