From spaces to places

Abstract

Mobile privacy concerns are central to Ubicomp and yet remain poorly understood. We advocate a diversified approach, enabling the cross-interpretation of data from complementary methods. However, mobility imposes a number of limitations on the methods that can be effectively employed. We discuss how we addressed this problem in an empirical study of mobile social networking. We report on how, by combining a variation of experience sampling and contextual interviews, we have started focusing on a notion of context in relation to privacy, which is subjectively defined by emerging socio-cultural knowledge, functions, relations and rules. With reference to Gieryn's sociological work, we call this place, as opposed to a notion of context that is objectively defined by physical and factual elements, which we call space. We propose that the former better describes the context for mobile privacy.

A machine-learning based approach to privacy-aware information-sharing in mobile social networks

“…By considering simple statistical models that take into account factors other than the geographic location, the authors show that the semantic category of the location being shared (such as a shopping center or a hospital) and the social group of the person asking for the location are significant factors in deciding whether to share the location. These results support earlier efforts [2,16,24] in providing a set of contextual features that have a statistically significant impact on the location-sharing behavior of mobile users. We use these results for our application when defining initial universal sharing policies, and we will describe them in the "Evaluation" section.…”

“…In the next question (16), the participants were asked to assign some of their Facebook friends to five different social categories (based on [34]): (1) school colleagues, (2) friends, (3) family members, (4) work colleagues and (5) acquaintances. Each participant could assign one Facebook contact to at most one category.…”

Quantified Baby

“…Each participant was asked to unpack the Owlet box and think aloud as they did so [49], to set up the device and use it for the first time (if possible) as the interaction was video recorded and observed. Immediately after their 'out-of-box' experience, participants started a combination elicitation [11] and feedback [44] diary study capturing media and making diary entries of use and non-use of the Owlet. A diary study offered a way to record the situation in context of every interaction with the Owlet, and it also ensured that participants recorded their experiences 'in the moment' [11].…”

Targeted Advertising on the Handset: Privacy and Security Challenges

“…For instance, a participant could be detected near a shopping centre or restaurant, and this might trigger questions about their advertising preferences. Experience Sampling has been combined successfully with mobile devices, e.g., (Mancini et al, 2009;Consolvo and Walker, 2003;Ben Abdesslem et al, 2010), but care should be taken before implementing a large-scale mobile advertising ESM study. For instance, the ethical, legal and privacy implications of collecting mobile location and advertising data need to be considered (Henderson and Ben Abdesslem, 2009), and mechanisms for anonymising data accordingly need to be designed, especially given the number of related datasets that have been subsequently deanonymised by researchers (Ohm, 2009).…”