Dynamic Skyline Queries on Encrypted Data Using Result Materialization
share
Skyline computation is an increasingly popular query, with broad applicability in domains such as healthcare, travel and finance. Given the recent trend to outsource databases and query evaluation, and due to the proprietary and sometimes highly sensitivity nature of the data (e.g., in healthcare), it is essential to evaluate skylines on encrypted datasets. Several research efforts acknowledged the importance of secure skyline computation, but existing solutions suffer from at least one of the following shortcomings: (i) they only provide ad-hoc security; (ii) they are prohibitively expensive; or (iii) they rely on unrealistic assumptions, such as the presence of multiple non-colluding parties in the protocol. Inspired from solutions for secure nearest-neighbors (NN) computation, we conjecture that the most secure and efficient way to compute skylines is through result materialization. However, this approach is significantly more challenging for skylines than for NN queries. We exhaustively study and provide algorithms for pre-computation of skyline results, and we perform an in-depth theoretical analysis of this process. We show that pre-computing results while minimizing storage overhead is NP-hard, and we provide dynamic programming and greedy heuristics that solve the problem more efficiently, while maintaining storage at reasonable levels. Our algorithms are novel and applicable to plain-text skyline computation, but we focus on the encrypted setting where materialization reduces the cost of skyline computation from hours to seconds. Extensive experiments show that we clearly outperform existing work in terms of performance, and our security analysis proves that we obtain a smaller (and quantifiable) data leakage than competitors.
READ FULL TEXT