Publication detail

NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned

MIN, S. FAJČÍK, M. DOČEKAL, M. ONDŘEJ, K. SMRŽ, P.

Original Title

NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned

Type

article in a collection out of WoS and Scopus

Language

English

Original Abstract

We review the EfficientQA competition from NeurIPS 2020. The competition focused on open-domain question answering (QA), where systems take natural language questions as input and return natural language answers. The aim of the competition was to build systems that can predict correct answers while also satisfying strict on-disk memory budgets. These memory budgets were designed to encourage contestants to explore the trade-off between storing retrieval corpora or the parameters of learned models. In this report, we describe the motivation and organization of the competition, review the best submissions, and analyze system predictions to inform a discussion of evaluation for open-domain QA.

Keywords

question answering, QA, ODQA, efficientQA, memory, disk memory, budget, efficient parameter, retrieval corpora

Authors

MIN, S.; FAJČÍK, M.; DOČEKAL, M.; ONDŘEJ, K.; SMRŽ, P.

Released

1. 8. 2021

Publisher

Proceedings of Machine Learning Research

Location

online

ISBN

2640-3498

Year of study

133

Number

133

Pages from

86

Pages to

111

Pages count

25

URL

http://proceedings.mlr.press/v133/min21a/min21a.pdf

BibTex

@inproceedings{BUT175821,
  author="MIN, S. and FAJČÍK, M. and DOČEKAL, M. and ONDŘEJ, K. and SMRŽ, P.",
  title="NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned",
  booktitle="Proceedings of the NeurIPS 2020 Competition and Demonstration Track",
  year="2021",
  series="Proceedings of Machine Learning Research",
  volume="133",
  number="133",
  pages="86--111",
  publisher="Proceedings of Machine Learning Research",
  address="online",
  issn="2640-3498",
  url="http://proceedings.mlr.press/v133/min21a/min21a.pdf"
}