Privacy-Preserving Vital Node Identification in Complex Networks: Evaluating Centrality Measures under Limited Network Information

Abstract

The identification of vital nodes in complex networks is of immense significance across a wide range of research domains and applications, including social network analysis, epidemiology, and physics. These fields frequently employ centrality measures to study complex networks. Despite the increasing emphasis on privacy, there has been limited research on its role in identifying vital nodes. This is particularly relevant in instances where networks contain sensitive information that is not widely shared, such as in Bluetooth-based contact networks.

Our study aims to evaluate the performance of centrality measures in determining node vitality under varying levels of privacy. Advanced algorithms for ranking the vitality of nodes are evaluated under the condition that only limited information about neighbors is available. Through comprehensive simulation experiments, we identify the most suitable algorithms for privacy-sensitive estimation of node vitality and shed light on the role of network traits and specific algorithm properties such as multi-aggregation. By bridging the gap between vital node identification and privacy considerations, this research contributes to a more comprehensive understanding of privacy-aware approaches in the study of complex networks.