Edge Thresholding Working Group Report

Abstract When working with very large networks it is typical for scientists to present a “thinned out ” version of the network in order to avoid the clutter of the entire network. For example in the hypothetical case of illustrating trading patterns between groups of nations it might be appropriate to limit the inclusion of inter-nation edges to all those that are significant in terms of their weight but do not, say, associate with a country outside the grouping. Arising from a discussion during one of the introductory sessions we became interested in a problem relating to the discovery of “key events ” in a network, in terms of an ordered addition of edges to the network.