Measuring And Eliminating Information Disclosure In Publishing Views Process

As an effective method for information exchange, view publishing provide convenience for data exchange and data sharing, but the problem of sensitive information disclosure in the view publishing process is more serious, and becomes a new research hotspot in database security. This paper analyzes the current situation of the domestic and international security problem in view publishing process, and researches for the problem of information disclosure from a completely new perspective.At first, the definition of sensitive information, the origin of sensitive information and the types of sensitive information are introduced. Some methods for eliminating information disclosure are introduced, with emphasis on the analysis of conditional probability independence method and k-anonymity protecting method. These methods decide view security before publishing views, and eliminate information disclosure about views existing information disclosure.Secondly, on the basis of the above research and existent solutions, depending on the idea of conditional probability, the deciding view security algorithm under prior knowledge is proposed. It can decide whether view set is security or not. The formula for measuring information disclosure is proposed. Using the idea of generalization, eliminating information disclosure algorithm based on Leak-Value is proposed. It can eliminate information disclosure about view set existing information disclosure effectively, and publish view set after eliminating information disclosure.Moreover, according to the idea of k-anonymity protecting method, k-anonymity protecting method based on association cover and k-anonymity protecting method based on quasi-identifier are proposed. These two methodsdecide view security from different angles. They eliminate information disclosure using the idea of generalization and suppression, and measure precision about view set after eliminating information disclosure based on precision measuring formula. At last, optimal view set is published.Finally, we analyze the view security deciding methods by instances, validate the algorithms using experiment, and prove the feasibility and validity of algorithms.