Management of Users'' Privacy Preferences in Context

摘要

There has been intensive research on user-controlled privacy from the perspective of agent automation of privacy related user tasks. The W3C''s Platform for Privacy Preferences (P3P) specifies standards that can be used by P3P agents to automatically retrieve a web-site''s privacy policy on how the users'' data is collected, stored, and managed, and then to determine whether they are compatible with the user''s privacy preferences. Current approaches to managing user''s privacy do not capture context, are not user-friendly, and do cater well to the dynamic nature of privacy preferences very well. Clearly, the user''s privacy preferences depend on the context of the user''s online activity and the user''s preferences evolve with user''s experience and changing levels of trust in various organizations and domains. We propose a model for user''s privacy preferences that incorporates the context for user activity and we apply it using a Case-based Reasoning (CBR) approach that relates the current activity to previous activities stored in the case-base and thus forms an intuitive and understandable process. We describe the context model for privacy preferences, how CBR is used to create a new contextual case from the web-site''s privacy policy and the user''s current activity, and how the CBR retrieves matching cases to be applied to the retrieved privacy policy.