Investigating Privacy Preservation and Data Mining Performance in Big Data: A Hybrid Approach Using K-Anonymity, Genetic Algorithms, and Simulated Annealing

Abstract

With the exponential growth of big data preventing privacy while maintaining data utility has become a significant challenge. This research investigates the impact of k-anonymity on the performance of data mining classifiers in big data environments, focusing on accuracy, precision, and recall. Furthermore, a hybrid optimization model leveraging Genetic Algorithms (GA) and Simulated Annealing (SA) is proposed to balance privacy preservation and information loss effectively. Results demonstrate that hybrid techniques can optimize k-anonymity parameters, ensuring robust privacy without significant degradation of classification performance.