Um conjunto de técnicas de inspeção orientado à avaliação de privacidade em redes sociais online

Abstract

The growing discussion on user privacy in Online Social Networks (OSNs) has encouraged the adoption of good design and evaluation practices to ensure the social acceptability and quality of use of these applications. In this way, privacy in an OSN has become one of the determining factors of quality, since discrepant privacy mechanisms can negatively influence the adoption of these applications by the user. Currently, several researchers have been presenting technologies to address privacy issues in OSNs. Some of the technologies present alternative solutions to be applied at design time. Another part of these technologies provides access controls and models for managing personal information. However, these technologies are not intended to detect real problems that could compromise user interaction with the privacy scenario provided by these applications. An alternative way of supporting privacy assessment to detect potential problems is through inspection. Inspection methods stand out because they are interesting not only because they are inexpensive, but also because they are adequate to guarantee the quality of use. In this sense, we propose a set of evaluation techniques oriented to privacy inspection in the context of OSN. The proposed techniques were developed from a systematic mapping of the literature that pointed to the lack of technologies that would support the privacy inspection in the environment in question. The techniques were evaluated and evolved through experimental studies that identified their feasibility as inspection techniques. The results indicate that the use of the techniques is potentially useful for the detection of real privacy defects due to their exploratory and reflexive nature. In this way, we hoped to contribute to promote the improvement of the quality of privacy in social applications, so that the set of techniques can be used to ensure the availability of the necessary functionalities and at the same time guarantee the privacy of the user data.