Towards Identifying the Best New Connection in a Spatial Network: Optimising the Performance of Hole Discovery

摘要

Networks are used to represent phenomena such that we can measure their structure. Spatial networks are a special class of networks that reflect the embedding space within which the network is contained, incorporating the property of spatial autocorrelation and its impact on network measures. These measures of spatial network structure have thus far largely focused on the structure of the network, as opposed to the absence of that structure. This paper builds on past work in identifying an aspect of the absence of structure, that of identifying holes or chordless cycles in a network. It is the first step in identifying the best new connection to make in the network, identifying where the absence of structure is having the most significant impact. This paper presents the implementation of optimisations in discovering network holes.