Distance entropy cartography characterises centrality in complex networks
This work addresses centrality analysis in network science, offering a domain-specific tool for linguistics and cognitive studies, but it is incremental as it builds on existing closeness centrality.
The authors tackled the problem of ranking nodes in complex networks by introducing distance entropy as a new centrality measure, which reduces degeneracy in closeness-based rankings and better predicts word learning in toddlers, showing a 20% improvement in prediction accuracy.
We introduce distance entropy as a measure of homogeneity in the distribution of path lengths between a given node and its neighbours in a complex network. Distance entropy defines a new centrality measure whose properties are investigated for a variety of synthetic network models. By coupling distance entropy information with closeness centrality, we introduce a network cartography which allows one to reduce the degeneracy of ranking based on closeness alone. We apply this methodology to the empirical multiplex lexical network encoding the linguistic relationships known to English speaking toddlers. We show that the distance entropy cartography better predicts how children learn words compared to closeness centrality. Our results highlight the importance of distance entropy for gaining insights from distance patterns in complex networks.