| The rapid development of computer and information technology has brought about the era of big data,which has facilitated people’s life and information exchange,but also brought about the challenge of privacy security.The civil aviation industry is an industry with high concentration of privacy information,and there are many privacy leaks in the industry,and the civil aviation industry is currently in a critical transformation stage,so the study of civil aviation data information security has not only important theoretical research significance,but also practical application significance.The thesis focuses on the study of differential privacy-based data dissemination algorithms and the application of differential privacy algorithms to single-dimensional data and multi-dimensional data,taking into account the data characteristics of civil aviation data.The thesis firstly analyzes the data characteristics of civil aviation single-dimensional data,adopts the histogram publishing method to protect data privacy according to its data characteristics,and evaluates the algorithm results functionally;secondly,to address the shortcomings of differential privacy algorithm in multi-dimensional data publishing,an algorithm is selected that can improve the usability of data by adding noise to more sensitive and more important data attributes to be protected,and evaluates the algorithm results functionally.The algorithm results are evaluated in terms of functionality.The experimental results show that the algorithm used in this thesis is feasible.Finally,a prototype civil aviation differential privacy data distribution system is designed and implemented based on a non-interactive framework where data processors can process and distribute data,and data analysts can query and download data from the system. |
