Adaptive Improvement Algorithm for Big Data Privacy Information Entropy under Differential Privacy and k-Anonymity

The theory of differential privacy and kanonymity aims to provide a solution to the privacy leakage problem for the collection and abuse of massive data in the era of big data. We propose an improved privacy protection algorithm based on this. For differential privacy, the Gaussian mechanism dynamically adjusts privacy budgets to accelerate data processing without compromising privacy. For kanonymity, an information entropy-based partitioning method optimizes equivalence classes, improving anonymized data utility while reducing computational costs. Validated on medical big data, the algorithm demonstrates superior privacy protection, processing efficiency, and data usability, offering a balanced privacy-utility solution with theoretical and practical significance for data management.