Graph sampling for node embedding

TL;DR

This paper introduces sampling methods for node embedding in graphs, enhancing computational efficiency and scalability by leveraging eigenvectors of the graph Laplacian and node features.

Contribution

It proposes novel sampling approaches that improve efficiency in node embedding, addressing scalability issues in graph representation learning.

Findings

Sampling methods achieve comparable accuracy with full-graph methods

Enhanced scalability for large graphs

Effective use of Laplacian eigenvectors and node features

Abstract

Node embedding is a central topic in graph representation learning. Computational efficiency and scalability can be challenging to any method that requires full-graph operations. We propose sampling approaches to node embedding, with or without explicit modelling of the feature vector, which aim to extract useful information from both the eigenvectors related to the graph Laplacien and the given values associated with the graph.