Reducing side effects of hiding sensitive itemsets in privacy preserving data mining.

Area of Science:

• Computer Science
• Data Science
• Information Security

Background:

• Data mining extracts knowledge from large datasets.
• Protecting private or confidential data is crucial.
• Privacy-preserving data mining (PPDM) addresses data security concerns.

Purpose of the Study:

• To propose a novel algorithm for privacy-preserving data mining.
• To hide sensitive information within datasets.
• To minimize the negative impacts of data sanitization.

Main Methods:

• Introduced the Hiding-Missing-Artificial Utility (HMAU) algorithm.
• Implemented transaction deletion strategy based on sensitive-to-non-sensitive ratios.
• Incorporated weights for hiding failures, missing itemsets, and artificial itemsets.

Main Results:

• The HMAU algorithm effectively hides sensitive itemsets.
• Evaluated performance based on execution time, deleted transactions, and side effects.
• Demonstrated the algorithm's ability to manage privacy-utility trade-offs.

Conclusions:

• The HMAU algorithm offers a novel approach to PPDM.
• User-defined weights allow customization of privacy-utility balance.
• The method provides a practical solution for sanitizing sensitive data.