Privacy preserving database application testing
Xintao Wu, Yongge Wang, Yuliang Zheng · 2003
Traditionally, application software developers carry out their tests on their own local development databases. However, such local databases usually have only a small number of sample data and hence cannot simulate satisfactorily a live environment, especially in terms of performance and scalability testing. On the other hand, the idea of testing applications over live production databases is increasingly problematic in most situations primarily due to the fact that such use of live production databases has the potential to expose sensitive data to an unauthorized tester and to incorrectly update information in the underlying database. In this paper, we investigate techniques to generate mock databases for application software testing without revealing any confidential information from the live production databases. Specifically, we will design mechanisms to create the deterministic rule set R, non-deterministic rule set N R, and statistic data set S for a live production database. We will then build a security Analyzer which will process the triplet together with security requirements (security policy) and output a new triplet The security Analyzer will guarantee that no confidential information could be inferred from the new triplet The mock database generated from this new triplet can simulate the live environment for testing purpose, while maintaining the privacy of data in the original database.