Network Members Can Hide from Group Centrality Measures

Group centrality measures are a generalization of standard centrality, designed to quantify the importance of not just a single node (as is the case with standard measures) but rather that of a group of nodes. Some nodes may have an incentive to evade such measures, i.e., to hide their actual import...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2023-12
Main Authors: Waniek, Marcin, Rahwan, Talal
Format: Article
Language:English
Subjects:
Network topologies
Nodes
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Group centrality measures are a generalization of standard centrality, designed to quantify the importance of not just a single node (as is the case with standard measures) but rather that of a group of nodes. Some nodes may have an incentive to evade such measures, i.e., to hide their actual importance, in order to conceal their true role in the network. A number of studies have been proposed in the literature to understand how nodes can rewire the network in order to evade standard centrality, but no study has focused on group centrality to date. We close this gap by analyzing four group centrality measures: degree, closeness, betweenness, and GED-walk. We show that an optimal way to rewire the network can be computed efficiently given the former measure, but the problem is NP-complete given closeness and betweenness. Moreover, we empirically evaluate a number of hiding strategies, and show that an optimal way to hide from degree group centrality is also effective in practice against the other measures. Altogether, our results suggest that it is possible to hide from group centrality measures based solely on the local information available to the group members about the network topology.
ISSN:2331-8422