Toward Practical Privacy-Preserving Frequent Itemset Mining on Encrypted Cloud Data

Shuo Qiu, Boyang Wang, Ming Li, Jiqiang Liu, Yanfeng Shi

Research output: Contribution to journalArticlepeer-review

13 Scopus citations


Frequent itemset mining, which is the essential operation in association rule mining, is one of the most widely used data mining techniques on massive datasets nowadays. With the dramatic increase on the scale of datasets collected and stored with cloud services in recent years, it is promising to carry this computation-intensive mining process in the cloud. Amount of work also transferred the approximate mining computation into the exact computation, where such methods not only improve the accuracy also aim to enhance the efficiency. However, while mining data stored on public clouds, it inevitably introduces privacy concerns on sensitive datasets. In this paper, we propose a new framework for enforcing privacy in frequent itemset mining, where data are both collected and mined in an encrypted form in a public cloud service. We specifically design three secure frequent itemset mining protocols on top of this framework. Our first protocol achieves more efficient mining performance while our second protocol provides a stronger privacy guarantee. In order to further optimize the performance of the second protocol, we leverage a minor trade-off of privacy to get our third protocol. Finally, we evaluate the performance of our protocols with extensive experiments, and the results demonstrate that our protocols obviously outperform previous solutions in performance with the same security level.

Original languageEnglish (US)
Article number8010376
Pages (from-to)312-323
Number of pages12
JournalIEEE Transactions on Cloud Computing
Issue number1
StatePublished - Jan 1 2020
Externally publishedYes


  • Cloud computing
  • efficiency
  • frequent itemset mining
  • privacy-preserving

ASJC Scopus subject areas

  • Software
  • Information Systems
  • Hardware and Architecture
  • Computer Science Applications
  • Computer Networks and Communications


Dive into the research topics of 'Toward Practical Privacy-Preserving Frequent Itemset Mining on Encrypted Cloud Data'. Together they form a unique fingerprint.

Cite this