Privacy preserving record linkage in the presence of missing values

作者：

Highlights：

• It is proposed that the missing value in a record is handled by utilising the values of the corresponding fields in the k-NNs of this record.

• The proposed method for dealing with missing values allows the use of the traditional blocking techniques to handle the scalability issue.

• The existing Bloom filter protocol has been adapted to address both issues of missing values and privacy preservation.

摘要

•It is proposed that the missing value in a record is handled by utilising the values of the corresponding fields in the k-NNs of this record.•The proposed method for dealing with missing values allows the use of the traditional blocking techniques to handle the scalability issue.•The existing Bloom filter protocol has been adapted to address both issues of missing values and privacy preservation.

论文关键词：Record linkage,Probabilistic record linkage,Privacy preserving record linkage,Missing values,k-Nearest Neighbours,Data encryption

论文评审过程：Received 24 October 2016, Revised 19 June 2017, Accepted 5 July 2017, Available online 5 July 2017, Version of Record 5 September 2017.

论文官网地址：https://doi.org/10.1016/j.is.2017.07.001