January 17, 2023

by Noah Lloyd

Finding the shortest path in a traditional network shouldn’t be an issue, and “is a straightforward task when the network of interest is fully known,” but what happens when the size of large networks means their maps “are substantially incomplete”? The authors of “Finding shortest and nearly shortest path nodes in large substantially incomplete networks by hyperbolic mapping” have determined that large, real-world networks “are not random but are organized according to latent-geometric rules.”

This being the case, they have discovered that, when “mapped to points in latent hyperbolic spaces,” they can calculate the shortest paths “along geodesic curves connecting endpoint nodes.” In other words, they can get from A to C, without knowing where B is.

This research has potential bearing on real-world large networks, from “cellular pathway reconstruction [to] communication security.”

Read “Finding shortest and nearly shortest path nodes in large substantially incomplete networks by hyperbolic mapping” and see the full list of authors below.

View on Site

Faculty: Maksim Kitsak

Topics: Network ScienceNetworks