F-Classify: Fuzzy Rule Based Classification Method for Privacy Preservation of Multiple Sensitive Attributes.

Area of Science:

• Computer Science
• Data Privacy
• Artificial Intelligence

Background:

• The rapid growth of data in smart environments necessitates robust privacy preservation techniques.
• Anonymizing datasets with multiple sensitive attributes is challenging due to attribute correlations.
• Existing methods often struggle with the complexity of multiple sensitive attributes.

Purpose of the Study:

• To propose a novel fuzzy logic-based privacy model for anonymizing datasets with multiple sensitive attributes.
• To address the unique challenges posed by attribute correlation in multi-attribute anonymization.
• To enhance both data privacy and utility in the context of sensitive information.

Main Methods:

• Development of the F-Classify model utilizing fuzzy logic for attribute classification.
• Classification of quasi-identifier and multiple sensitive attributes based on defined rules.
• Verification of the F-Classify algorithm's functionality using Hierarchical Label Propagation Network (HLPN).

Main Results:

• F-Classify demonstrates superior performance in privacy and data utility compared to existing methods.
• Experimental validation on healthcare datasets confirms the model's effectiveness.
• The artificial intelligence-based approach offers a lower execution time.

Conclusions:

• F-Classify provides an effective solution for privacy preservation in datasets with multiple sensitive attributes.
• Fuzzy logic offers a promising approach for tackling complex data anonymization challenges.
• The model balances privacy protection with data utility, making it suitable for real-world applications.