Record linkage allows data from different sources to be integrated to facilitate data mining tasks. However, in many cases, records have to be linked by personally identifiable information. To prevent privacy breaches, ideally records should be linked in a private way such that no information other than the matching result is leaked in the process. In this paper, we present an exact Private Record Linkage (PRL) protocol and an approximate PRL protocol. The exact PRL protocol is based on Oblivious Bloom Intersection, which is an efficient private set intersection protocol. The approximate PRL protocol extends the exact PRL protocol by incorporating Locality Sensitive Hash functions. Both protocols are secure in the semi-honest model. We also report the evaluation results based on our C implementation of the protocols. The results show that our protocols are efficient and effective.
|Title of host publication||Proceedings of 29th ACM Symposium on Applied Computing|
|Place of Publication||New York|
|Number of pages||7|
|Publication status||Published - 2014|
- record linkage study
- private record linkage
- data mining