@inproceedings{thawani-etal-2021-numeracy,
title = "Numeracy enhances the Literacy of Language Models",
author = "Thawani, Avijit and
Pujara, Jay and
Ilievski, Filip",
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2021",
address = "Online and Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.emnlp-main.557",
doi = "10.18653/v1/2021.emnlp-main.557",
pages = "6960--6967",
abstract = "Specialized number representations in NLP have shown improvements on numerical reasoning tasks like arithmetic word problems and masked number prediction. But humans also use numeracy to make better sense of world concepts, e.g., you can seat 5 people in your {`}room{'} but not 500. Does a better grasp of numbers improve a model{'}s understanding of other concepts and words? This paper studies the effect of using six different number encoders on the task of masked word prediction (MWP), as a proxy for evaluating literacy. To support this investigation, we develop Wiki-Convert, a 900,000 sentence dataset annotated with numbers and units, to avoid conflating nominal and ordinal number occurrences. We find a significant improvement in MWP for sentences containing numbers, that exponent embeddings are the best number encoders, yielding over 2 points jump in prediction accuracy over a BERT baseline, and that these enhanced literacy skills also generalize to contexts without annotated numbers. We release all code at https://git.io/JuZXn.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="thawani-etal-2021-numeracy">
<titleInfo>
<title>Numeracy enhances the Literacy of Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Avijit</namePart>
<namePart type="family">Thawani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jay</namePart>
<namePart type="family">Pujara</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Filip</namePart>
<namePart type="family">Ilievski</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online and Punta Cana, Dominican Republic</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Specialized number representations in NLP have shown improvements on numerical reasoning tasks like arithmetic word problems and masked number prediction. But humans also use numeracy to make better sense of world concepts, e.g., you can seat 5 people in your ‘room’ but not 500. Does a better grasp of numbers improve a model’s understanding of other concepts and words? This paper studies the effect of using six different number encoders on the task of masked word prediction (MWP), as a proxy for evaluating literacy. To support this investigation, we develop Wiki-Convert, a 900,000 sentence dataset annotated with numbers and units, to avoid conflating nominal and ordinal number occurrences. We find a significant improvement in MWP for sentences containing numbers, that exponent embeddings are the best number encoders, yielding over 2 points jump in prediction accuracy over a BERT baseline, and that these enhanced literacy skills also generalize to contexts without annotated numbers. We release all code at https://git.io/JuZXn.</abstract>
<identifier type="citekey">thawani-etal-2021-numeracy</identifier>
<identifier type="doi">10.18653/v1/2021.emnlp-main.557</identifier>
<location>
<url>https://aclanthology.org/2021.emnlp-main.557</url>
</location>
<part>
<date>2021-11</date>
<extent unit="page">
<start>6960</start>
<end>6967</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Numeracy enhances the Literacy of Language Models
%A Thawani, Avijit
%A Pujara, Jay
%A Ilievski, Filip
%S Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
%D 2021
%8 November
%I Association for Computational Linguistics
%C Online and Punta Cana, Dominican Republic
%F thawani-etal-2021-numeracy
%X Specialized number representations in NLP have shown improvements on numerical reasoning tasks like arithmetic word problems and masked number prediction. But humans also use numeracy to make better sense of world concepts, e.g., you can seat 5 people in your ‘room’ but not 500. Does a better grasp of numbers improve a model’s understanding of other concepts and words? This paper studies the effect of using six different number encoders on the task of masked word prediction (MWP), as a proxy for evaluating literacy. To support this investigation, we develop Wiki-Convert, a 900,000 sentence dataset annotated with numbers and units, to avoid conflating nominal and ordinal number occurrences. We find a significant improvement in MWP for sentences containing numbers, that exponent embeddings are the best number encoders, yielding over 2 points jump in prediction accuracy over a BERT baseline, and that these enhanced literacy skills also generalize to contexts without annotated numbers. We release all code at https://git.io/JuZXn.
%R 10.18653/v1/2021.emnlp-main.557
%U https://aclanthology.org/2021.emnlp-main.557
%U https://doi.org/10.18653/v1/2021.emnlp-main.557
%P 6960-6967
Markdown (Informal)
[Numeracy enhances the Literacy of Language Models](https://aclanthology.org/2021.emnlp-main.557) (Thawani et al., EMNLP 2021)
ACL
- Avijit Thawani, Jay Pujara, and Filip Ilievski. 2021. Numeracy enhances the Literacy of Language Models. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 6960–6967, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics.