This study introduces a fast lossless graph compression algorithm based on modern hyperbolic embedders that beats state-of-the-art by up to 42% on real-world graphs.
Abstract
Network theoreticians hypothesize that the structure of real-world networks has a geometric origin. Especially, hyperbolic geometry was proven insightful in representing and modeling of scale-free networks. Embedders are algorithms used to find a geometric representation of a network. In this study, we introduce a fast lossless graph compression algorithm based on modern hyperbolic embedders. Experimental validation on real-world and generated networks shows that our algorithm beats state-of-the-art by up to 42% on real-world graphs.
This work benchmarks 13 unsupervised hyperbolic graph embedders under a unified protocol for link prediction and topology reconstruction on synthetic and empirical networks and identifies the network regimes in which different paradigms succeed or fail and provides practical guidance for method selection in downstream...
Robert Jankowski, M. Kitsak, Dorota Celińska-Kopczyńska· 0 citations
The findings illustrate that geometric insights grounded in hyperbolic geometry can offer powerful tools for understanding, embedding, and visualizing complex graph structures.
†. SalouaNaama, Kave Salamatian, M. Crovella· 0 citations
An algorithm is given that computes a minimum set of edges that makes a connected plane graph 2-edge-connected in \(O(|V|(1+\alpha(|V|)))\) time and linear space, where \(\alpha\) is the inverse Ackermann function.
K. Dehaleesan, Asif Khan, Pranabendu Misra· International Symposium on M...· 0 citations
We start by surveying known properties of first passage percolation (FPP) geodesics on a Gromov-hyperbolic group $G$. Due to a coalescence phenomenon established in earlier work of the authors, a random tree $T(\xi,\omega)$ consisting of infinite random geodesics to a point $\xi$ on the Gromov boundary $\partial G$ eme...
It is studied whether an observed graph can distinguish a random graph generated by latent geometry from one with independent edges, which is an Erdős–Rényi graph with the same edge density.
Cong-Ying Luo· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
