Premium
Secure similarity coefficients computation for binary data and its extensions
Author(s) -
Zhang Bo,
Zhang Fangguo
Publication year - 2013
Publication title -
concurrency and computation: practice and experience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.309
H-Index - 67
eISSN - 1532-0634
pISSN - 1532-0626
DOI - 10.1002/cpe.3059
Subject(s) - similarity (geometry) , computer science , protocol (science) , computation , secure two party computation , secure multi party computation , binary number , theoretical computer science , security parameter , cryptography , algorithm , mathematics , artificial intelligence , arithmetic , alternative medicine , pathology , medicine , image (mathematics)
SUMMARY Similarity measures play an important role in classification problems, cluster analysis, and identification issues. This paper studies the secure similarity coefficients computation in the two‐party setting. Recently, a privacy‐preserving similarity coefficients protocol for binary data was proposed by Wong and Kim (Computers and Mathematics with Application 2012). We point out that their protocol is not secure, even in the semi‐honest model. In their protocol, the client can retrieve the inputs of the server without deviating from the protocol. Next, we propose a secure similarity coefficients computation protocol in the presence of malicious adversaries, which solves the same similarity coefficients functionality as that proposed by Wong and Kim. Meanwhile, we prove the protocol secure against the malicious adversaries by using the standard simulation‐based security definitions for secure two‐party computation. Also several extensions of our protocol for settling other specific problems are discussed. At last, we present a protocol computing the similarity coefficients with better privacy by using the secure integer division on ciphertexts. Copyright © 2013 John Wiley & Sons, Ltd.