Reasoning with imprecise privacy preferences

Abstract

User reluctance and context-dependent factors during information disclosure imply that people cannot always be counted on to indicate their appropriate privacy preference. This phenomenon is the well-known 'privacy paradox', which shows that users of modern technologies are constantly concerned about their privacy, but do not apply these concerns to their usage behaviour accordingly. The problem is that this mismatch between privacy concerns and the indicated privacy preference in software, is not considered when reasoning about the satisfaction of privacy requirements. This paper is a research vision that draws connections between the imprecisions in user privacy preferences, and reasoning about the satisfaction of privacy requirements. We outline the close relationship between privacy and user beliefs and uncertainties. We then propose a multi-agent framework that leverage on this relationship when reasoning about the satisfaction of privacy requirements. We anticipate that this vision will help reduce the gap between an increasingly complex information age and the software techniques needed to protect user privacy.