Privacy agents in the IoT : considerations on how to balance agent autonomy and user control in privacy decisions

Abstract

This thesis explored aspects that can help balance the level of user control and system autonomy for intelligent privacy agents in the context of the Internet of Things. This proposed balance could be reached considering aspects related to wanting to be interrupted to have control and being able to be interrupted to exert this control. Through literature review of interruption and privacy literature, variables related to these two perspectives were identified. This led to the variable set "Intelligent Privacy Interruptions”. To verify and validate this set, two research actions were performed. The first one was an online survey that allowed us to perform a sanity check that these variables were acceptable in this work’s context. The second was an experience sampling user study with 21 participants that allowed us to better understand how user behavior is informed by these variables. Based on these two interventions it was possible to note that the selected variables seem to show relevance and that they can be used to inform the development and design of privacy agents. The limitations of the partial results notwithstanding, through a quantitative analysis of data collected form the user study and the qualitative analysis of the exit interviews, it was possible to note a common mental process between the participants of the user study when deciding whether to withhold or delegate decision control to the agent. Future studies should be performed to verify the possibility of expansion and creation of a behavior and preference model that can be integrated to the decisionmaking system of intelligent privacy agents.