Cargando…

Bridging the Gap of AutoGraph Between Academia and Industry: Analyzing AutoGraph Challenge at KDD Cup 2020

Graph structured data is ubiquitous in daily life and scientific areas and has attracted increasing attention. Graph Neural Networks (GNNs) have been proved to be effective in modeling graph structured data and many variants of GNN architectures have been proposed. However, much human effort is ofte...

Descripción completa

Detalles Bibliográficos
Autores principales: Xu, Zhen, Wei, Lanning, Zhao, Huan, Ying, Rex, Yao, Quanming, Tu, Wei-Wei, Guyon, Isabelle
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Frontiers Media S.A. 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9245425/
https://www.ncbi.nlm.nih.gov/pubmed/35783353
http://dx.doi.org/10.3389/frai.2022.905104
_version_ 1784738736575086592
author Xu, Zhen
Wei, Lanning
Zhao, Huan
Ying, Rex
Yao, Quanming
Tu, Wei-Wei
Guyon, Isabelle
author_facet Xu, Zhen
Wei, Lanning
Zhao, Huan
Ying, Rex
Yao, Quanming
Tu, Wei-Wei
Guyon, Isabelle
author_sort Xu, Zhen
collection PubMed
description Graph structured data is ubiquitous in daily life and scientific areas and has attracted increasing attention. Graph Neural Networks (GNNs) have been proved to be effective in modeling graph structured data and many variants of GNN architectures have been proposed. However, much human effort is often needed to tune the architecture depending on different datasets. Researchers naturally adopt Automated Machine Learning on Graph Learning, aiming to reduce human effort and achieve generally top-performing GNNs, but their methods focus more on the architecture search. To understand GNN practitioners' automated solutions, we organized AutoGraph Challenge at KDD Cup 2020, emphasizing automated graph neural networks for node classification. We received top solutions, especially from industrial technology companies like Meituan, Alibaba, and Twitter, which are already open sourced on GitHub. After detailed comparisons with solutions from academia, we quantify the gaps between academia and industry on modeling scope, effectiveness, and efficiency, and show that (1) academic AutoML for Graph solutions focus on GNN architecture search while industrial solutions, especially the winning ones in the KDD Cup, tend to obtain an overall solution (2) with only neural architecture search, academic solutions achieve on average 97.3% accuracy of industrial solutions (3) academic solutions are cheap to obtain with several GPU hours while industrial solutions take a few months' labors. Academic solutions also contain much fewer parameters.
format Online
Article
Text
id pubmed-9245425
institution National Center for Biotechnology Information
language English
publishDate 2022
publisher Frontiers Media S.A.
record_format MEDLINE/PubMed
spelling pubmed-92454252022-07-01 Bridging the Gap of AutoGraph Between Academia and Industry: Analyzing AutoGraph Challenge at KDD Cup 2020 Xu, Zhen Wei, Lanning Zhao, Huan Ying, Rex Yao, Quanming Tu, Wei-Wei Guyon, Isabelle Front Artif Intell Artificial Intelligence Graph structured data is ubiquitous in daily life and scientific areas and has attracted increasing attention. Graph Neural Networks (GNNs) have been proved to be effective in modeling graph structured data and many variants of GNN architectures have been proposed. However, much human effort is often needed to tune the architecture depending on different datasets. Researchers naturally adopt Automated Machine Learning on Graph Learning, aiming to reduce human effort and achieve generally top-performing GNNs, but their methods focus more on the architecture search. To understand GNN practitioners' automated solutions, we organized AutoGraph Challenge at KDD Cup 2020, emphasizing automated graph neural networks for node classification. We received top solutions, especially from industrial technology companies like Meituan, Alibaba, and Twitter, which are already open sourced on GitHub. After detailed comparisons with solutions from academia, we quantify the gaps between academia and industry on modeling scope, effectiveness, and efficiency, and show that (1) academic AutoML for Graph solutions focus on GNN architecture search while industrial solutions, especially the winning ones in the KDD Cup, tend to obtain an overall solution (2) with only neural architecture search, academic solutions achieve on average 97.3% accuracy of industrial solutions (3) academic solutions are cheap to obtain with several GPU hours while industrial solutions take a few months' labors. Academic solutions also contain much fewer parameters. Frontiers Media S.A. 2022-06-16 /pmc/articles/PMC9245425/ /pubmed/35783353 http://dx.doi.org/10.3389/frai.2022.905104 Text en Copyright © 2022 Xu, Wei, Zhao, Ying, Yao, Tu and Guyon. https://creativecommons.org/licenses/by/4.0/This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
spellingShingle Artificial Intelligence
Xu, Zhen
Wei, Lanning
Zhao, Huan
Ying, Rex
Yao, Quanming
Tu, Wei-Wei
Guyon, Isabelle
Bridging the Gap of AutoGraph Between Academia and Industry: Analyzing AutoGraph Challenge at KDD Cup 2020
title Bridging the Gap of AutoGraph Between Academia and Industry: Analyzing AutoGraph Challenge at KDD Cup 2020
title_full Bridging the Gap of AutoGraph Between Academia and Industry: Analyzing AutoGraph Challenge at KDD Cup 2020
title_fullStr Bridging the Gap of AutoGraph Between Academia and Industry: Analyzing AutoGraph Challenge at KDD Cup 2020
title_full_unstemmed Bridging the Gap of AutoGraph Between Academia and Industry: Analyzing AutoGraph Challenge at KDD Cup 2020
title_short Bridging the Gap of AutoGraph Between Academia and Industry: Analyzing AutoGraph Challenge at KDD Cup 2020
title_sort bridging the gap of autograph between academia and industry: analyzing autograph challenge at kdd cup 2020
topic Artificial Intelligence
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9245425/
https://www.ncbi.nlm.nih.gov/pubmed/35783353
http://dx.doi.org/10.3389/frai.2022.905104
work_keys_str_mv AT xuzhen bridgingthegapofautographbetweenacademiaandindustryanalyzingautographchallengeatkddcup2020
AT weilanning bridgingthegapofautographbetweenacademiaandindustryanalyzingautographchallengeatkddcup2020
AT zhaohuan bridgingthegapofautographbetweenacademiaandindustryanalyzingautographchallengeatkddcup2020
AT yingrex bridgingthegapofautographbetweenacademiaandindustryanalyzingautographchallengeatkddcup2020
AT yaoquanming bridgingthegapofautographbetweenacademiaandindustryanalyzingautographchallengeatkddcup2020
AT tuweiwei bridgingthegapofautographbetweenacademiaandindustryanalyzingautographchallengeatkddcup2020
AT guyonisabelle bridgingthegapofautographbetweenacademiaandindustryanalyzingautographchallengeatkddcup2020