Mitigating Anonymity Challenges in Automated Testing and Debugging Systems

Modern software often provides automated testing and bug reporting facilities that enable developers to improve the software after release. Alas, this comes at the cost of user anonymity: reported execution traces may identify users. We present a way to mitigate this inherent tension between developer utility and user anonymity: automatically transform execution traces in a way that preserves their utility for testing and debugging while, at the same time, providing k-anonymity to users, i.e., a guarantee that the trace can at most identify the user as being part of a group of k indistinguishable users. We evaluate this approach in the context of an automated testing and bug reporting system for smartphone applications.