TWC SBE: Small: Protecting the Online Privacy of Users of Social Networks

人们普遍认为，保护网络隐私很重要，现代社会在许多方面都体现了这一关切。初步研究表明，第三方只需少量的爬行和计算资源，并使用简单的数据挖掘启发式算法，就可以潜在地将在线服务和公开可用的信息结合在一起，创建生活在任何目标地理区域的用户的详细资料。该研究调查了一些措施，这些措施可以显著改善用户的隐私保护，同时不会降低他们的整体互联网体验。重点是不太可信的第三方(例如，数据经纪人、广告商、垃圾邮件发送者、恶意软件分发者和恋童癖)，他们可以从许多不同的线上和线下来源收集、聚合和推断信息。这项研究有两个相互关联的研究推动力。首先，它探索第三方可以在多大程度上收集、汇总和统计处理来自OSN和其他在线和离线来源的信息，以创建配置文件。这一努力是开发严格的统计方法和概率模型，以估计潜在的隐私泄露程度。其次，这项研究调查了政府可以制定的各种隐私政策，以及OSN可以采取的广泛措施，以降低隐私风险。对于有希望的政策和措施组合，这项研究量化了隐私保护和可用性之间的权衡。

It is generally recognized that protecting online privacy is important, with modern society manifesting this concern in many ways. Preliminary research indicates that third parties, with modest crawling and computational resources, and employing simple data mining heuristics, can potentially combine online services and publicly available information to create detailed profiles of the users living in any targeted geographical area.This research investigates measures that can significantly improve privacy protection of users, while not degrading their overall Internet experience. The focus is on less-trustworthy third parties (e.g., data brokers, advertisers, spammers, malware distributors, and pedophiles), who can scrape, aggregate and infer information from many different online and offline sources. This research has two interrelated research thrusts. First, it explores to what extent third parties can collect, aggregate, and statistically process information from OSNs and other online and offline sources to create profiles. This thrust is developing rigorous statistical methodologies and probabilistic models for estimating the degree of potential privacy leakage. Second, this research investigates a variety of privacy policies that governments can establish, and a wide range of measures OSNs can take, to reduce the privacy risk. For promising combinations of policies and measures, this research quantifies the trade-off between privacy protection and usability.