Analysis of Privacy Patterns from An Architectural Perspective

Privacy has become an essential software quality that every software architect and developer must consider. Embedding privacy practices from the early stages of the system design is a necessary step to safeguard personal data from privacy violations. Privacy patterns are proposed in both industry and academia as reusable design solutions to tackle general privacy issues. However, these privacy patterns generally lack perspectives of software architecture thinking that would assist system developers to weave the patterns into concrete system designs. In addition, there is not yet a thorough investigation into the potential trade-offs amongst the software quality attributes when applying privacy patterns onto the systems. Privacy patterns have been reviewed from architectural contexts and analyzed with regards to the privacy properties and their impact on the performance. In this paper, we extend the trade-off analysis with modifiability. We then propose an enhanced pattern catalogue that shows the interrelation of privacy patterns and the system by organising privacy patterns according to their implementation space, covering architectural context and the impact analysis on different quality attributes. The information captured by the pattern catalogue provides a decision guide for software architects and developers when navigating the privacy pattern collection during system design phases.