Balancing privacy needs with location sharing in mobile computing

Mobile phones are increasingly becoming tools for social interaction. As more phones come equipped with location tracking capabilities, capable of collecting and distributing personal information (including location) of their users, user control of location information and privacy for that matter, has become an important research issue. This research first explores various techniques of user control of location in location-based systems, and proposes the re-conceptualisation of deception (defined here as the deliberate withholding of location information) from information systems security to the field of location privacy. Previous work in this area considers techniques such as anonymisation, encryption, cloaking and blurring, among others. Since mobile devices have become social tools, this thesis takes a different approach by empirically investigating first the likelihood of the use of the proposed technique (deception) in protecting location privacy. We present empirical results (based on an online study) that show that people are willing to deliberately withhold their location information to protect their location privacy. However, our study shows that people feel uneasy in engaging in this type of deception if they believe this will be detected by their intended recipients. The results also suggest that the technique is popular in situations where it is very difficult to detect that there has been a deliberate withholding of location information during a location disclosure. Our findings are then presented in the form of initial design guidelines for the design of deception to control location privacy. Based on these initial guidelines, we propose and build a deception-based privacy control model. Two different evaluation approaches are employed in investigating the suitability of the model. These include; a field-based study of the techniques employed in the model and a laboratory-based usability study of the Mobile Client application upon which the DPC model is based, using HCI (Human Computer Interaction) professionals. Finally, we present guidelines for the design of deception in location disclosure, and lessons learned from the two evaluation approaches. We also propose a unified privacy preference framework implemented on the application layer of the mobile platform as a future direction of this thesis