项目名称: 数据流发布中的隐私保护理论和方法研究
项目编号: No.61502111
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 自动化技术、计算机技术
项目作者: 王金艳
作者单位: 广西师范大学
项目金额: 20万元
中文摘要: 由于数据共享的紧迫需求以及公众对隐私问题的担忧,数据发布中的隐私保护问题已成为数据库与信息安全领域交叉的研究热点。数据流具有海量性、实时性和动态变化性,这对传统的针对静态数据集发布的隐私保护模型、数据可用性度量方法和隐私保护方法提出了严重的挑战。本项目针对关系型数据流和高维集值型数据流发布中的隐私保护问题展开研究。首先利用流推理语言等分别针对关系型数据流和集值型数据流对攻击者的背景知识进行建模,提出相应的隐私保护模型,同时利用集合论、模糊数等综合考虑数据的信息损失和时延提出数据可用性度量方法。在此基础上,利用熵理论、进化算法、模糊集理论、软集理论等针对关系型数据流和集值型数据流的发布分别提出相应的隐私保护方法。项目的研究将为数据流发布提供系统的隐私保护理论和方法,对经济发展、社会稳定以及互联网技术的有效利用具有重要的推动作用。
中文关键词: 数据发布;隐私保护;隐私模型;数据可用性;背景知识
英文摘要: Privacy preserving data publishing is a hot research topic in the crossover field between databases and information security, because of the urgent requirement for information sharing and the public fear for privacy leakage. Data streams are massive, real-time and volatile, and the privacy protection models, measure of data utility and privacy preserving techniques for static data publishing cannot be applied on streaming data. The project researches on the privacy preserving problem for publishing relational data streams and high-dimensional set-valued data streams. Firstly, we model the background knowledge of attacks by using stream reasoning language, and present privacy protection models in relational data streams and set-valued data streams, respectively. Also, we analyze data information loss and time delay by using set theory, fuzzy number, etc., to give the method to measure the data utility. Furthermore, we utilize well-developed theories such as entropy theory, evolutionary computation, fuzzy set theory, soft set theory, to design privacy preserving techniques for publishing relational data streams and set-valued data streams, respectively. The project will offer systematic theories and methods for privacy preserving data streams publishing, and promote economic developments, social stability and the efficient use of internet technology.
英文关键词: Data publishing;Privacy preserving;Privacy model;Data utility;Background knowledge