Everything is Connected: Graph Neural Networks

  • 2023-01-19 18:09:43
  • Petar Veličković
  • 32

Abstract

In many ways, graphs are the main modality of data we receive from nature.This is due to the fact that most of the patterns we see, both in natural andartificial systems, are elegantly representable using the language of graphstructures. Prominent examples include molecules (represented as graphs ofatoms and bonds), social networks and transportation networks. This potentialhas already been seen by key scientific and industrial groups, withalready-impacted application areas including traffic forecasting, drugdiscovery, social network analysis and recommender systems. Further, some ofthe most successful domains of application for machine learning in previousyears -- images, text and speech processing -- can be seen as special cases ofgraph representation learning, and consequently there has been significantexchange of information between these areas. The main aim of this short surveyis to enable the reader to assimilate the key concepts in the area, andposition graph representation learning in a proper context with related fields.

 

Quick Read (beta)

loading the full paper ...