Authorization-aware optimization for multi-provider queries
Ekaterina B. Dimitrova, Panos K. Chrysanthis, Adam J. Lee · 2019
The sharing of sensitive personal information via cloud platforms motivates the need for measures that aim to minimize information leakage to unauthorized users. In this work, we propose a novel SQL optimizer that strikes a balance between query runtime performance and private information exposure. Our approach to ensuring that the access control policies regulating data disclosure are enforced during distributed query execution is based upon a state-of-the-art authorization model from the literature and a preference-aware query optimizer. Our preliminary studies show that our approach outperforms the way the authorization model was originally implemented in terms of query runtime performance which is crucial for the operation on Big Data. To improve it, we adjust the algorithms utilized by a preference-aware query optimizer.