Um modelo de design de privacidade para o compartilhamento de informações pessoais em redes sociais online

Abstract

In the last few years, Social Network Sites (SNSs) have experienced a growth in their number of participants, becoming an increasingly embedded part of people's daily lives. However, along with such growth, the possibility of closer interactions between people, brought by these systems, has triggered users' concerns and issues regarding privacy, given the increasingly risk of people have their personal information being improperly accessed within such environments. Thus, the idea of privacy by design, i.e., incorporating privacy principles at design time rather than as an afterthought, is presented as an essential requirement, even if challenging, in order to create user interfaces that allow users to express their privacy requirements regarding personal information disclosure. Thereby, in this thesis, we intent to provide designers with a better understanding of how privacy regarding personal information disclosure can be addressed in SNSs. In this direction, we present the Privacy Design Model (PDM) -- a descriptive model that is built upon Altman's theory of privacy and Semiotic Engineering and considers as personal information in SNSs not only pieces of information about the individual, but also the individual's speech and activities that express, direct or indirectly, his/her opinions and views within the system. PDM aims at helping designers to take into consideration how different aspects influence the level of privacy being offered to users, by structuring the design space of personal information disclosure in SNSs through a number of dimensions. The evaluation of PDM expressivity shows that it is expressive enough to represent relevant aspects of privacy and also descriptive enough to express differences in the privacy models of SNSs with distinct purposes. The potential epistemic value of MDP is shown, by fostering discussions and reflections regarding privacy decisions that could be useful for the design of SNSs' privacy models.