Model-Based Analysis of Privacy in Electronic Services

Many stakeholders are involved in complex electronic services. Individuals release personal information to front-ends in advanced web services. However, to what extent the acquired information is distributed and processed, what other data is included and/or merged in user profiles, remains unclear. Much more personal information is often released than strictly necessary. As is well known, many authentication technologies release personal data and make transactions linkable. Service providers can accurately profile individuals and trade profiles with external commercial entities. Moreover, service providers delegate their responsibilities for the data they collected to external entities who receive these data, while consequently, individual users lose control of their personal information. The present dissertation presents a logic based modeling approach to inspect privacy in composite electronic services from different viewpoints in data protection. The first part provides the background for the major concepts on privacy complemented by an overview of the different privacy analysis approaches currently available. The second part presents a logic based modeling approach for inspecting the user privacy. First of all, the key concepts necessary to reason on privacy in composite electronic services are identified. Examples are privacy policies, authentication technologies; but also, the privacy preferences and trust assumptions of users. These concepts are mapped to components of a logic based framework supporting automated privacy inspection of composite electronic services. The approach extracts profiles, containing a set of personal data that can be compiled by service providers after a user has interacted with a particular service provider. The resulting profiles constitute the input for a qualitative privacy analysis capable of providing meaningful feedback to both end-users and service designers. The former can select a service based on the minimal data disclosure principle. The latter can use the framework to select privacy-enhancing technologies and define consistent privacy policies from the earliest design stage. In the third part, the approach is validated on two electronic public transport ticketing systems. The case study shows the extensibility of the framework. The final part focuses on service provider accountability, one of the core privacy protection principles of the upcoming European General Data Protection Regulation. Service provider accountability refers to the obligation of data controllers to demonstrate compliance of their data practices with the rules and regulations. In this part, an inference model providing individuals with global accountability guarantees of multi-component systems is introduced. It considers different user expectations on provided evidence (i.e. log evidence) of declared data handling practices. The feedback provided by this approach is useful for both end-users and auditors who conduct privacy assessments on behalf of the end-users.status: publishe