Secure Hamming distance based record linkage with malicious adversaries

•We present an secure protocol for computing the Hamming distance based record linkage with malicious adversaries.•We present protocols for computing the powers of an encrypted value and implementing zero test on an encrypted value.•We give out an extension to secure Levenshtein distance based record linkage.

Record linkage aims at finding the matching records from two or multiple different databases. Many approximate string matching methods in privacy-preserving record linkage have been presented. In this paper, we study the problem of secure record linkage between two data files in the two-party computation setting. We note that two records are linked when all the Hamming distances between their attributes are smaller than some fixed thresholds. We firstly construct two efficient secure protocols for computing the powers of an encrypted value and implementing zero test on an encrypted value, then present an efficient protocol within constant rounds for computing the Hamming distance based record linkage in the presence of malicious adversaries by transferring these two protocols. We also discuss the extension of our protocol for settling the Levenshtein distance based record linkage problem.

Graphical abstractFigure optionsDownload full-size imageDownload as PowerPoint slide