Scalable privacy-preserving t-repetition protocol with distributed medical data

Ji Young Chun, Dowon Hong, Dong Hoon Lee, Ik Rae Jeong

Research output: Contribution to journalArticlepeer-review

Abstract

Finding rare cases with medical data is important when hospitals or research institutes want to identify rare diseases. To extract meaningful information from a large amount of sensitive medical data, privacy-preserving data mining techniques can be used. A privacypreserving t-repetition protocol can be used to find rare cases with distributed medical data. A privacy-preserving t-repetition protocol is to find elements which exactly t parties out of n parties have in common in their datasets without revealing their private datasets. A privacy-preserving trepetition protocol can be used to find not only common cases with a high t but also rare cases with a low t. In 2011, Chun et al. suggested the generic set operation protocol which can be used to find t-repeated elements. In the paper, we first show that the Chun et al.'s protocol becomes infeasible for calculating t-repeated elements if the number of users is getting bigger. That is, the computational and communicational complexities of the Chun et al.'s protocol in calculating t-repeated elements grow exponentially as the number of users grows. Then, we suggest a polynomial-time protocol with respect to the number of users, which calculates t-repeated elements between users.

Original languageEnglish
Pages (from-to)2451-2460
Number of pages10
JournalIEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
VolumeE95-A
Issue number12
DOIs
Publication statusPublished - 2012 Dec

Keywords

  • Data mining
  • Privacy
  • Rare cases
  • Set operation
  • T-repetition

ASJC Scopus subject areas

  • Signal Processing
  • Computer Graphics and Computer-Aided Design
  • Electrical and Electronic Engineering
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'Scalable privacy-preserving t-repetition protocol with distributed medical data'. Together they form a unique fingerprint.

Cite this