Mapping global dynamics of benchmark creation and saturation in artificial intelligence

Recent studies raised concerns over the state of AI benchmarking, reporting issues such as benchmark overfitting, benchmark saturation and increasing centralization of benchmark dataset creation. To facilitate monitoring of the health of the AI benchmarking ecosystem, the authors introduce methodolo...

Full description

Bibliographic Details
Main Authors: Simon Ott, Adriano Barbosa-Silva, Kathrin Blagec, Jan Brauner, Matthias Samwald
Format: Article
Language:English
Published: Nature Portfolio 2022-11-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-022-34591-0
_version_ 1798019129027854336
author Simon Ott
Adriano Barbosa-Silva
Kathrin Blagec
Jan Brauner
Matthias Samwald
author_facet Simon Ott
Adriano Barbosa-Silva
Kathrin Blagec
Jan Brauner
Matthias Samwald
author_sort Simon Ott
collection DOAJ
description Recent studies raised concerns over the state of AI benchmarking, reporting issues such as benchmark overfitting, benchmark saturation and increasing centralization of benchmark dataset creation. To facilitate monitoring of the health of the AI benchmarking ecosystem, the authors introduce methodologies for creating condensed maps of the global dynamics of benchmark.
first_indexed 2024-04-11T16:35:42Z
format Article
id doaj.art-357a20a54edd4dfabde61f00141c7e58
institution Directory Open Access Journal
issn 2041-1723
language English
last_indexed 2024-04-11T16:35:42Z
publishDate 2022-11-01
publisher Nature Portfolio
record_format Article
series Nature Communications
spelling doaj.art-357a20a54edd4dfabde61f00141c7e582022-12-22T04:13:51ZengNature PortfolioNature Communications2041-17232022-11-0113111110.1038/s41467-022-34591-0Mapping global dynamics of benchmark creation and saturation in artificial intelligenceSimon Ott0Adriano Barbosa-Silva1Kathrin Blagec2Jan Brauner3Matthias Samwald4Institute of Artificial Intelligence, Medical University of Vienna. Währingerstraße 25aInstitute of Artificial Intelligence, Medical University of Vienna. Währingerstraße 25aInstitute of Artificial Intelligence, Medical University of Vienna. Währingerstraße 25aOxford Applied and Theoretical Machine Learning (OATML) Group, Department of Computer Science, University of OxfordInstitute of Artificial Intelligence, Medical University of Vienna. Währingerstraße 25aRecent studies raised concerns over the state of AI benchmarking, reporting issues such as benchmark overfitting, benchmark saturation and increasing centralization of benchmark dataset creation. To facilitate monitoring of the health of the AI benchmarking ecosystem, the authors introduce methodologies for creating condensed maps of the global dynamics of benchmark.https://doi.org/10.1038/s41467-022-34591-0
spellingShingle Simon Ott
Adriano Barbosa-Silva
Kathrin Blagec
Jan Brauner
Matthias Samwald
Mapping global dynamics of benchmark creation and saturation in artificial intelligence
Nature Communications
title Mapping global dynamics of benchmark creation and saturation in artificial intelligence
title_full Mapping global dynamics of benchmark creation and saturation in artificial intelligence
title_fullStr Mapping global dynamics of benchmark creation and saturation in artificial intelligence
title_full_unstemmed Mapping global dynamics of benchmark creation and saturation in artificial intelligence
title_short Mapping global dynamics of benchmark creation and saturation in artificial intelligence
title_sort mapping global dynamics of benchmark creation and saturation in artificial intelligence
url https://doi.org/10.1038/s41467-022-34591-0
work_keys_str_mv AT simonott mappingglobaldynamicsofbenchmarkcreationandsaturationinartificialintelligence
AT adrianobarbosasilva mappingglobaldynamicsofbenchmarkcreationandsaturationinartificialintelligence
AT kathrinblagec mappingglobaldynamicsofbenchmarkcreationandsaturationinartificialintelligence
AT janbrauner mappingglobaldynamicsofbenchmarkcreationandsaturationinartificialintelligence
AT matthiassamwald mappingglobaldynamicsofbenchmarkcreationandsaturationinartificialintelligence