Homogeneous network publishing privacy protection based on differential privacy uncertainty

With heterogeneous networks become more prevalent, there is a growing need for privacy pro-tection mechanisms in homogeneous networks within heterogeneous networks. Homogeneous networks are particularly vulnerable to malicious attacks when they are published. To solve this issue, we propose a privacy protection method for homogeneous network publishing based on differential privacy uncertainty. Our method is inspired by partitioning, which divides the ho-mogeneous network into community subgraphs and bridging subgraphs. Then, we use differential privacy and random perturbations to achieve the coding rearrangement and decoding of the subgraph uncertain trees. Next, we perturb the node degree sequence based on the generalization idea and design different reconstruction strategies to perturb the subgraph structure features. Finally, we perform postprocessing to fuse the subgraphs by measuring the sensitivity of the bridging nodes. We conducted extensive experimental analysis on real datasets, and the results show that our privacy protection method can simultaneously guarantee the privacy and avail-ability of homogeneous network data publishing.