Client-side multi-agent system for contextual private data management for e-commerce and m-commerce on the semantic web

现有的私人数据管理和相关的隐私增强工具（PETs）旨在增强用户对其私人信息的控制，但它们是碎片化的，功能和可用性有限，因此远远不适合大规模用户采用。根据许多国家最近的法律准则，还迫切需要客户端技术来补充并加强企业技术的目标，企业技术是组织雇员用来保护和管理客户私人数据的技术。目前，基于技术的隐私解决方案采取了一种“一刀切”的方法。本研究的目标是根据用户被要求发布私人信息的上下文，提供围绕用户决策的高级定制。上下文和自定义是重要而有用的机制，因为与数据处理有关的隐私、道德、社会和经济问题因用户配置文件、客户和其他利益相关方群体而异。这种关注也根据物理（空间和时间）情况和数据类型而有所不同。本研究提出了一个客户端私有数据管理系统，以支持用户在互联网和移动网络上进行商业活动时向组织发布上下文和隐私意识数据。将研究的科学问题包括合适的社会技术隐私架构、支持电子商务和移动商务中隐私感知用户上下文的知识表示、支持这些上下文的基于智能的算法和代理、用于提高隐私软件可用性的智能算法和个人代理，以及基于代理的新型隐私Web服务和基础设施。虽然本研究的目标是解决科学问题，但用户可见的结果是使用户能够在向组织发布私人数据时做出明智且高度细致的上下文决策的技术。这些技术还将提醒用户注意潜在的隐私问题，提醒他们进行电子商务或移动商务交易时的社会经济背景，并允许用户组织和查看他们在特定情况下向组织提供的私人数据的报告。

Existing private data management and related privacy enhancing tools (PETs), aimed at increasing the users' control over their private information, are fragmented, limited in capability and usability, and thus are far from suitable for wide scale user adoption. There is also an urgent need for client-side technologies to complement, and thus strengthen the goals, of the enterprise technologies that organizations' employees use to protect and manage customers' private data, in accordance to recent legal guidelines in many countries. Currently, technology-based privacy solutions take a one-size-fits-all approach. This research's goals are to provide a high-level of customization around user decision-making depending on the contexts in which the user is being asked to release private information. Contexts and customization are important and useful mechanisms, as privacy, ethical, social, and economic concerns regarding data handling differ according to user profiles, customer, and other stakeholder groups. Such concerns differ too according to physical (spatial and temporal) situation and types of data. This research proposes a client-side private data management system to support users' contextual and privacy-aware data release to organizations when conducting commerce over the Internet and mobile networks.  Scientific problems that will be investigated include suitable socio-technical privacy architecture, knowledge representations to support privacy-aware user contexts in e-commerce and m-commerce, intelligence-based algorithms and agents for supporting these contexts, intelligent algorithms and personal agents for increasing privacy software usability, and novel agent-based privacy Web services and infrastructure. While the objectives of this research address scientific problems, the user-visible outcomes are technologies that enable users to make informed and highly-nuanced contextual decisions around releasing private data to organizations. The technologies will also alert users as to potential privacy problems, remind them of socio-economic contexts when conducting e- or m-commerce transactions, and allow users to organize and view reports on the private data they have given to organizations under particular circumstances.