Bursa Technical University, Turkey
Centrality analysis is a common approach to identify influence of a node on other nodes in social networks. Based on the application, it may be more desirable to remove the most/least influential node from the network. Consider an outbreak of an infectious disease where an agent emerges, evolves, and spreads over several countries. It is crucial to take countermeasures such as vaccination, culling, social distancing, etc. and limit the scope of the consequences. To analyze the available data, a transmission graph can be defined where each transmission is represented by a directed edge from the source node to the destination node. In order to evaluate the importance of the nodes in the transmission graph, various centrality methods can be applied such as page rank, closeness centrality, betweenness centrality, etc. To avoid limiting results by only considering the shortest paths, we defined a novel centrality metric called adapted betweenness which also considers paths that are longer up to a user provided threshold. To apply countermeasure, we prefer more influential nodes in the network. On the other hand, consider a mobile sensor network partitioned into multiple disjoint segments. To restore network connectivity, a reactive recovery approach is restructuring the network through node mobility. However, movement of a node may cause further partitioning in the partition. Thus, it is crucial to identify nodes for movement such that the total movement cost can be minimized. The idea in this approach is identifying the least influential node among the candidates for movement and selecting it to perform the movement.