Advanced Security and Privacy Techniques for Secure Big Data Query, Sharing and Processing

信息通信技术的广泛应用，不断推动各领域数据爆发式增长。据估计，每天大约产生2.5万亿字节的数据，这些数据在决策、业务规划、知识发现等方面发挥着至关重要的作用。显然，这种情况导致了对大数据分析的持续兴趣。然而，大数据在给我们带来许多机遇的同时，也带来了新的挑战，特别是安全和隐私方面的挑战。如果我们不重视大数据安全，虚假注入的数据会让大数据驱动的应用变得毫无用处。另一方面，大多数有价值的数据通常都是个人和敏感的，如果我们不关注大数据隐私，数据所有者就会对共享他们的数据没有信心。因此，为了让加拿大适应大数据时代，大数据安全和隐私永远不应该是事后才想到的。相反，加拿大不仅应该培养精通大数据的新一代数据工程师，而且应该培养可靠和值得信赖的平台，以确保大数据时代安全和保护隐私的数据捕获、管理、存储、搜索和共享。提出的研究旨在解决大数据时代的安全和隐私挑战，特别是考虑到复杂的大数据查询、共享和处理中的安全和隐私威胁，这些威胁在以前报道的研究中尚未得到充分利用。这项研究的主要目标是通过使用一种跨学科的方法来研究一套高级安全和隐私技术，即结合密码学、高级数据结构和数据挖掘技术来保护各种大数据应用中的数据查询、共享和处理。特别是，这项建议将在以下四个方面解决大数据的“4V”(容量、速度、多样性和准确性)特性带来的重大技术挑战：i)开发高效且保护隐私的基于相似度的查询技术，以平衡eHealthcare大数据系统中的效用、隐私和效率；ii)开发高效且保护隐私的“天际线变体”查询技术，以满足各种真实场景的需求；iii)设计加密图表上的隐私保护查询技术，以保护访问模式隐私；以及iv)开发可靠的、隐私保护的、访问可控的框架，以确保大数据数字孪生系统中的数据共享和处理。这项拟议的研究将深入和广泛地利用首席调查员的研究专长，并得到新不伦瑞克大学计算机科学系的大力支持。前沿研究将为大数据安全的演变产生新的想法和知识，使HQP培训成为可能，并为加拿大提供新的安全和隐私保护的大数据解决方案。

The widespread application of information and communication technology has continuously promoted the explosive growth of data in various fields. It has been estimated that approximately 2.5 quintillion bytes of data are produced each day, and such data play a vital role in decision making, business planning, knowledge discovery, etc. Obviously, this situation has resulted in continuing interest in big data analytics. Nevertheless, while big data brings us many opportunities, it also introduces new challenges, especially security and privacy challenges. If we do not pay attention to big data security, false injected data would make big data driven applications useless. On the other hand, most valuable data are usually personal and sensitive, if we do not pay attention to big data privacy, data owners will have no confidence in sharing their data. Therefore, in order to adapt Canada to the big data era, big data security and privacy should never be an afterthought. Instead, Canada should prepare not only a new generation of data engineers skilled in big data, but also reliable and trustworthy platforms to ensure secure and privacy-preserving data capture, curation, storage, search, and sharing in big data era. The proposed research is envisioned to address security and privacy challenges in big data era, particularly considering the security and privacy threats in complex big data query, sharing, and processing, which have not yet been fully exploited in previously reported studies. The main objective of this proposed research is to investigate a set of advanced security and privacy techniques by using an interdisciplinary approach, i.e., combining cryptography, advanced data structures, and data mining techniques, to secure data query, sharing, and processing in various big data applications. In particular, this proposal will address significant technical challenges arising from "4V" (Volume, Velocity, Variety, and Veracity) characteristics of big data, in the following four thrusts: i) develop efficient and privacy-preserving similarity-based query techniques to balance utility, privacy, and efficiency in eHealthcare big data systems; ii) develop efficient and privacy-preserving "skyline variants" query techniques to fit various real scenarios' needs; iii) design privacy-preserving query techniques over encrypted graphs for preserving access pattern privacy; and iv) develop reliable, privacy-preserving, and access controllable frameworks to secure data sharing and processing in big data digital twin systems. This proposed research will draw intensively and extensively on the research expertise of the Principal Investigator, as well as the strong supports from the Faculty of Computer Science, University of New Brunswick. The cutting-edge research will generate new ideas and knowledge for the evolution of big data security, enable HQP training and provide new secure and privacy-preserving big data solutions for Canada.