creator Qian, Jianwei 数据共享中的安全隐私 2019 Fall 2019 Dissertation Illinois Institute of Technology CS / Computer Science advisor Wan, Peng-Jun advisor Li, Xiang-Yang Computer science Information technology Data privacy Data security Data sharing Open data Privacy and security Security and privacy en Data is a non-exclusive resource and has synergistic effects. Open data sharing will enhance the utilization of big data’s value and tremendously boost economic growth and transparency. Data sharing platforms have emerged worldwide, but with very limited services. Security is one of the main reasons why most data are not commonly shared. This dissertation aims to tackle several security issues in building a trustworthy data sharing ecosystem. First, I reveal the privacy risks in data sharing by designing de-anonymization and privacy inference attacks. Second, I present an analysis of the relationship between the attacker's knowledge and the privacy risk of data sharing, and try quantifying and estimating the risk. Then, I propose anonymization algorithms to protect the privacy of participants in data sharing. Finally, I survey the status quo, privacy and security concerns, and opportunities in data trading. This dissertation involves various data types with a focus on graph data and speech data; it also involves various forms of data sharing including collection, publishing, query, and trading.数据是一种非排他性的资源而且具有协同效应，开放数据共享将提升大数据价值的利用率并极大地促进经济增长和政企工作透明性。数据共享平台已在全世界出现但提供的服务仍然相当有限，安全隐私是数据共享还不普遍的主要原因之一。这篇毕业论文旨在为建立一个可信安全的数据共享生态系统而解决几个安全隐私问题。首先，我们设计去匿名和隐私信息推断攻击，以此为例揭示数据共享中的隐私问题；其次，我们量化数据共享的隐私风险和攻击者背景知识，并分析这两者之间的关系；然后，我们设计一些匿名算法来保护数据共享参与者的隐私；最后，我们对数据共享现状，各环节的安全隐私问题，现有技术和未来研究方向做综述。这篇论文涉及多种数据类型，重点关注图数据和语音数据，也涉及多种不同类型的数据分享，包括数据收集，数据发布，数据查询计算，数据交易。 born digital application/pdf In Copyright http://rightsstatements.org/page/InC/1.0/ Restricted Access Embargo Until: 02/11/2021 http://hdl.handle.net/10560/islandora:1009767