Measuring Apps' Privacy-Friendliness : Introducing transparency to apps' data access behavior

Mobile apps brought unprecedented convenience to everyday life, and nowadays, hardly any interactive service exists without having an interface through an app. The rich functionalities of apps rely on the pervasive capabilities of the mobile device, such as its cameras and other types of sensors. Consequently, apps generate a diverse and large amount of data, which can often be deemed as privacy-sensitive data. As the mobile device is also equipped with several means to transmit the collected data, such as WiFi and 4G, it brings further concerns about individuals' privacy. Even though mobile operating systems use access control mechanisms to guard system resources and sensors, apps exercise their granted privileges in an opaque manner. Depending on the type of privilege, apps require explicit approval from the user in order to acquire access to them through permissions. Nonetheless, granting permission does not put constraints on the access frequency. Granted privileges allow the app to access users' personal data for a long period of time, typically until the user explicitly revokes the access. Furthermore, available control tools lack monitoring features, and therefore, the user faces hindrances to comprehend the magnitude of personal data access. Such circumstances can erode intervenability from the interface of the phone, lead to incomprehensible handling of personal data, and thus, create privacy risks for the user. This thesis covers a long-term investigation of apps' data access behavior and makes an effort to shed light on various privacy implications. It also shows that app behavior analysis yields information that has the potential to increase transparency, to enhance privacy protection, to raise awareness regarding consequences of data disclosure, and to assist the user in informed decision-making while selecting apps or services. We introduce models, methods, and demonstrate the data disclosure risks with experimental results. Finally, we show how to communicate privacy risks through the user interface by taking the results of app behavior analyses into account.Mobile apps brought unprecedented convenience to everyday life, and nowadays, hardly any interactive service exists without having an interface through an app. The rich functionalities of apps rely on the pervasive capabilities of the mobile device. Consequently, apps generate a diverse and large amount of data, which can often be deemed as privacy-sensitive data. Even though mobile operating systems use access control mechanisms to guard system resources and sensors, apps exercise their granted privileges in an opaque manner. Furthermore, available control tools lack monitoring features, and therefore, the user faces hindrances to comprehend the magnitude of personal data access. This thesis covers a long-term investigation of apps' data access behavior and makes an effort to shed light on various privacy implications. It also shows that app behavior analysis yields information that has the potential to increase transparency, to enhance privacy protection, to raise awareness regarding consequences of data disclosure, and to assist the user in informed decision-making while selecting apps or services