Dynamic multidimensional data privacy protection algorithm

The current algorithms for K-anonymity are inadequate for clustering multi-dimensional data and cannot adapt to changes in anonymized datasets. To address these challenges, this paper introduces a Dynamic Multi-dimensional Data Privacy Protection Algorithm (DMPPA) that leverages enhancements in Mongoose optimization techniques. Initially, DMPPA employs a novel approach to cluster personal data using an advanced Mongoose optimization strategy. Furthermore, DMPPA enhances the differentiation of sensitive data and minimizes information loss rate, facilitating highly accurate clustering of multi-dimensional data. It also introduces a new method for dividing equivalence classes based on the quantity of information loss and the measurement of sensitive data differential values. Experimental results show that DMPPA surpasses existing methods such as MSAA(Multi-Stage Adaptive Algorithm), EPKM(Enhanced Pseudo K-Means Algorithm), and AECA(Adaptive Enhanced Clustering Algorithm) by enhancing clustering accuracy, increasing the average sensitivity of data differentiation, and decreasing the information loss rate.