Efficient Protocols for Privacy Preserving Matching Against Distributed Datasets

When datasets are distributed on different sources, finding out matched data while preserving the privacy of the datasets is a widely required task. In this paper, we address two matching problems against the private datasets on N (N ≥ 2) parties. The first one is the Privacy Preserving Set Intersection (PPSI) problem, in which each party wants to learn the intersection of the N private datasets. The second one is the Privacy Preserving Set Matching (PPSM) problem, in which each party wants to learn whether its elements can be matched in any private set of the other parties. For the two problems we propose efficient protocols based on a threshold cryptosystem which is additive homomorphic. In a comparison with the related work in [18], the computation and communication costs of our PPSI protocol decrease by 81% and 17% respectively, and the computation and communication costs of our PPSM protocol decrease by 80% and 50% respectively. In practical utilities both of our protocols save computation time and communication bandwidth.