Skip to main navigation Skip to search Skip to main content

Probing Negative Sampling for Contrastive Learning to Learn Graph Representations

  • Shiyi Chen
  • , Ziao Wang
  • , Xinni Zhang
  • , Xiaofeng Zhang*
  • , Dan Peng
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Graph representation learning has long been an important yet challenging task for various real-world applications. However, its downstream tasks are mainly performed in the settings of supervised or semi-supervised learning. Inspired by recent advances in unsupervised contrastive learning, this paper is thus motivated to investigate how the node-wise contrastive learning could be performed. Particularly, we respectively resolve the class collision issue and the imbalanced negative data distribution issue. Extensive experiments are performed on three real-world datasets and the proposed approach achieves the SOTA model performance.

Original languageEnglish
Title of host publicationMachine Learning and Knowledge Discovery in Databases. Research Track - European Conference, ECML PKDD 2021, Proceedings
EditorsNuria Oliver, Fernando Pérez-Cruz, Stefan Kramer, Jesse Read, Jose A. Lozano
PublisherSpringer Science and Business Media Deutschland GmbH
Pages434-449
Number of pages16
ISBN (Print)9783030865191
DOIs
StatePublished - 2021
Externally publishedYes
Event21st Joint European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2021 - Virtual, Online
Duration: 13 Sep 202117 Sep 2021

Publication series

NameLecture Notes in Computer Science
Volume12976 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference21st Joint European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2021
CityVirtual, Online
Period13/09/2117/09/21

Keywords

  • Contrastive learning
  • Graph neural network
  • Negative sampling

Fingerprint

Dive into the research topics of 'Probing Negative Sampling for Contrastive Learning to Learn Graph Representations'. Together they form a unique fingerprint.

Cite this