A Framework for Enforcing Privacy in Mining Frequent Patterns

• Author(s) / Creator(s)

  • Oliveira, Stanley
  • Zaiane, Osmar

• Technical report TR02-13. Discovering hidden patterns from large amounts of data plays an important role in marketing, business, medical analysis, and other applications where these patterns are paramount for strategic decision making. However, recent research has shown that some discovered patterns can pose a threat to security and privacy. One alternative to address the privacy requirements in mining hidden patterns is to look for a balance between hiding restrictive patterns and disclosing non-restrictive ones. In this paper, we propose a new framework for enforcing privacy in mining frequent itemsets. We combine, in a single framework, three techniques for efficiently hiding restrictive patterns: a transaction retrieval engine relying on inverted files and boolean queries; and a set of algorithms to ``sanitize'' a database. In addition, we introduce a mining association performance measure that quantifies the fraction of mining patterns which are preserved after sanitizing a database. We also report the results of performance evaluation of our research prototype and an analysis of such results. | TRID-ID TR02-13

• Date created

  2002

• Subjects / Keywords

  • Data Mining and Privacy
  • Database Systems
  • Artifactual Patterns
  • Hiding Failure
  • Sanitizing Algorithms
  • Misses Cost

• Type of Item

  Report

• DOI

  https://doi.org/10.7939/R3RF5KJ5P

• License

  Attribution 3.0 International