English

Paths and flows for centrality measures in networks

Combinatorics 2021-02-09 v2

Abstract

We consider the number of paths that must pass through a subset XX of vertices of a network NN in a maximum sequence of arc-disjoint paths connecting two vertices yy and zz. We show that when XX is a singleton, that number equals the difference between the maximum flow value from yy to zz in NN and the maximum flow value from yy to zz in the network obtained by NN setting to zero the capacities of arcs incident to XX. That fact theoretically justifies the common identification of those two concepts in network literature. We also show that the same equality does not hold when X2.|X|\geq 2. Consequently, two conceptually different group centrality measures involving paths and flows can naturally be defined, both extending the classic flow betweenness centrality.

Keywords

Cite

@article{arxiv.2003.13338,
  title  = {Paths and flows for centrality measures in networks},
  author = {Daniela Bubboloni and Michele Gori},
  journal= {arXiv preprint arXiv:2003.13338},
  year   = {2021}
}
R2 v1 2026-06-23T14:31:39.032Z