Fast link prediction for large networks using spectral embedding

TL;DR

This paper introduces scalable spectral embedding-based link prediction algorithms that are efficient for large networks, maintaining accuracy while significantly reducing computation time.

Contribution

The paper presents a novel class of link prediction algorithms using spectral embedding and the k closest pairs algorithm, enabling scalability to very large networks.

Findings

Achieve comparable accuracy to standard algorithms

Significantly faster runtime on large networks

Effective for large-scale link prediction tasks

Abstract

Many link prediction algorithms require the computation of a similarity metric on each vertex pair, which is quadratic in the number of vertices and infeasible for large networks. We develop a class of link prediction algorithms based on a spectral embedding and the k closest pairs algorithm that are scalable to very large networks. We compare the prediction accuracy and runtime of these methods to existing algorithms on several large link prediction tasks. Our methods achieve comparable accuracy to standard algorithms but are significantly faster.