Edge Importance in Complex Networks

Silvia Noschese,Lothar Reichel

2024-07-05

Abstract:Complex networks are made up of vertices and edges. The latter connect the vertices. There are several ways to measure the importance of the vertices, e.g., by counting the number of edges that start or end at each vertex, or by using the subgraph centrality of the vertices. It is more difficult to assess the importance of the edges. One approach is to consider the line graph associated with the given network and determine the importance of the vertices of the line graph, but this is fairly complicated except for small networks. This paper compares two approaches to estimate the importance of edges of medium-sized to large networks. One approach computes partial derivatives of the total communicability of the weights of the edges, where a partial derivative of large magnitude indicates that the corresponding edge may be important. Our second approach computes the Perron sensitivity of the edges. A high sensitivity signals that the edge may be important. The performance of these methods and some computational aspects are discussed. Applications of interest include to determine whether a network can be replaced by a network with fewer edges with about the same communicability.

Physics and Society,Social and Information Networks

What problem does this paper attempt to address?

The problem that this paper attempts to address is how to evaluate the importance of edges in complex networks. Specifically, the authors explore two methods to estimate the importance of edges in medium to large-scale networks:

1. **Partial Derivative Method**: This method assesses the importance of an edge by calculating the partial derivative of the total communication volume with respect to the edge weight. If the partial derivative of a particular edge is relatively large, it indicates that this edge might be important.
2. **Perron Sensitivity Method**: This method evaluates the importance of an edge by calculating its Perron sensitivity. High sensitivity indicates that the edge might be important.

The paper also discusses the performance and computational issues of these methods in practical applications, such as how to determine whether a network can maintain approximately the same communication capacity by reducing the number of edges. These issues have significant applications in many fields, including road network optimization, social network analysis, and more.