Hiding emerging patterns with local recoding generalization

Michael W.K. Cheng, Koon Kau CHOI, Kwok Wai CHEUNG

Research output: Chapter in book/report/conference proceedingConference proceedingpeer-review

4 Citations (Scopus)

Abstract

Establishing strategic partnership often requires organizations to publish and share meaningful data to support collaborative business activities. An equally important concern for them is to protect sensitive patterns like unique emerging sales opportunities embedded in their data. In this paper, we contribute to the area of data sanitization by introducing an optimization-based local recoding methodology to hide emerging patterns from a dataset but with the underlying frequent itemsets preserved as far as possible. We propose a novel heuristic solution that captures the unique properties of hiding EPs to carry out iterative local recoding generalization. Also, we propose a metric which measures (i) frequentitemset distortion that quantifies the quality of published data and (ii) the degree of reduction in emerging patterns, to guide a bottom-up recoding process. We have implemented our proposed solution and experimentally verified its effectiveness with a benchmark dataset.

Original languageEnglish
Title of host publicationAdvances in Knowledge Discovery and Data Mining - 14th Pacific-Asia Conference, PAKDD 2010, Proceedings
Pages158-170
Number of pages13
EditionPART 1
DOIs
Publication statusPublished - 2010
Event14th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2010 - Hyderabad, India
Duration: 21 Jun 201024 Jun 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume6118 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference14th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2010
Country/TerritoryIndia
CityHyderabad
Period21/06/1024/06/10

Scopus Subject Areas

  • Theoretical Computer Science
  • Computer Science(all)

User-Defined Keywords

  • Data sanitization
  • Emerging patterns
  • Frequent itemsets
  • Pattern hiding

Fingerprint

Dive into the research topics of 'Hiding emerging patterns with local recoding generalization'. Together they form a unique fingerprint.

Cite this