III: Small: Collaborative Research: Privacy-Aware Collaborative Data Sharing in Human-Centered Social Networks

在在线社交网络(OSN)的帮助下，人们分享个人和公共信息，并与朋友、同事、同事、家人甚至陌生人建立社交联系。因此，OSN存储了大量关于用户及其交互的敏感信息。为了保护这些信息，隐私控制一直被视为OSN的一个核心特征。尽管OSN目前提供了一些隐私控制机制，允许用户控制对他们共享的信息的访问，但不幸的是，用户无法控制其他人共享的数据。例如，如果用户在朋友的空间中发布评论，S/他无法指定哪些用户可以查看该评论。同样，当用户上传一张照片并标记照片中出现的朋友时，标记的朋友不能限制谁可以看到这张照片。由于多个关联用户可能对共享数据有不同的隐私担忧，隐私冲突会发生，缺乏协作隐私控制增加了朋友向公众泄露敏感信息的潜在风险。为了解决这样一个关键问题，必须适应来自多个关联用户的不同隐私控制要求，以协作管理OSN中的共享数据。该项目开发的新技术将极大地提高OSN中隐私感知数据共享的最新水平，并将对未来OSN中的协作共享系统的设计产生影响。此外，由于电子健康记录和金融数据共享等许多其他协作环境中的隐私做法也需要多个用户共同管理信息隐私，因此该项目产生的基本成果可以扩展到社交网络以外的协作环境。本项目的目标是寻求一种有效和灵活的机制来实现OSN中的隐私感知协同数据共享。为此，研究人员将首先分析与OSN中多个用户相关的数据共享，并阐明几种典型的隐私冲突场景，以了解这些冲突带来的风险。为了减少隐私冲突带来的风险，研究人员将研究一种协作数据共享机制，以支持多个隐私问题的规范和执行。此外，将建立一个系统的冲突检测和解决机制，并评估其处理在开放式网络中数据共享的协作管理中发生的隐私冲突的能力。该项目中的冲突解决试图通过对隐私风险和共享损失的定量分析来平衡隐私保护的需要和用户共享信息的愿望。提出的方法的另一个引人注目的特点是对OSN中隐私控制的理论和经验分析的支持。研究人员将使用博弈论模型分析理性用户的策略行为，其中每个参与者旨在通过调整OSN中协作数据共享的隐私设置来尽可能地满足她/他的隐私关切。此外，研究人员还将对实际用户行为进行实证分析，并将其与协作数据共享中的理论分析进行对比，阐明理论模型与实际用户行为之间的差距。有关这个项目的更多细节，包括实验数据和课程材料，可以在项目网站(www.cs.clemson.edu/~hongxih/project/gpc)上找到。

With the help of Online Social Networks (OSNs), people share personal and public information and make social connections with friends, coworkers, colleagues, family and even with strangers. As a result, OSNs store a huge amount of sensitive information about users and their interactions. To protect such information, privacy control has been treated as a central feature of OSNs. Although OSNs currently provide some privacy control mechanisms allowing users to regulate access to information they share, users, unfortunately, have no control over data others share. For instance, if a user posts a comment in a friend's space, s/he cannot specify which users can view the comment. Similarly, when a user uploads a photo and tags friends who appear in the photo, the tagged friends cannot restrict who can see this photo. Since multiple associated users may have different privacy concerns over the shared data, privacy conflicts occur and the lack of collaborative privacy control increases the potential risk of leaking sensitive information by friends to the public. To address such a critical issue, it is essential to accommodate diverse privacy control requirements coming from multiple associated users for collaboratively managing the shared data in OSNs. The new techniques developed in this project will substantially enhance the state-of-the-art in privacy-aware data sharing in OSNs and will have implications for the design of future collaborative sharing systems in OSNs. Moreover, since privacy practices in many other collaborative environments, such as electronic health records and financial data sharing also require multiple users to co-manage the privacy of information, the fundamental results generated by this project could be expanded to those collaborative environments beyond social networks. The goal of this project is to seek an effective and flexible mechanism to enable privacy-aware collaborative data sharing in OSNs. To this end, the researchers will first analyze data sharing associated with multiple users in OSNs, and articulate several typical scenarios of privacy conflicts to understand the risks posed by those conflicts. To mitigate risks caused by privacy conflicts, the researchers will investigate a collaborative data sharing mechanism to support the specification and enforcement of multiple privacy concerns. In addition, a systematic conflict detection and resolution mechanism will be created and evaluated with respect to its ability to cope with privacy conflicts occurring in collaborative management of data sharing in OSNs. The conflict resolution in the project attempts to balance the need for privacy protection and users' desire for information sharing by quantitative analysis of privacy risk and sharing loss. Another compelling feature of the proposed approach is the support of both theoretical and empirical analyses on privacy control in OSNs. The researchers will analyze the strategic behaviors of rational users using a game-theoretic model, where each player aims at accommodating her/his privacy concerns as much as possible by adjusting her/his privacy setting in collaborative data sharing in OSNs. Furthermore, the researchers will carry out empirical analysis for practical user behaviors and contrast it with the theoretical analysis in collaborative data sharing, articulating the gap between the theoretic model and real user behaviors. More details about this project, including experimental data and curricular materials can be found on the project website (www.cs.clemson.edu/~hongxih/projects/gpc).