Network Robustness Assessment via Edge Criticality Evaluation: Improvement of Bridgeness and Topological Overlap Methods by the Iterative Metrics Re-estimation

Detection of significant edges maintaining the connectivity in complex networks is essential in many applications such as attack vulnerability analysis, the spread of epidemic diseases, and information spreading patterns discovery. There are many existing methods enabling us to evaluate the criticality ranking of links in networks, which are based on straightforward algorithms and topological features of analyzed graphs. In this paper, we propose an improvement of well-known algorithms for the edge criticality ranking in complex networks, bridgeness, and topological overlap. The two resulting methods, iterative bridgeness, and iterative topological overlap are both funded on the principle of iterative metrics re-estimation. To compare methods, we conducted a numerical simulation and decomposition of three real-world benchmark networks. The resulting algorithms are compared with the original methods and demonstrated an increase of efficiency of up to 40% for the iterative bridgeness and up to 11% for the iterative topological overlap.