@inproceedings{ma-etal-2021-issues,
title = "Issues with Entailment-based Zero-shot Text Classification",
author = "Ma, Tingting and
Yao, Jin-Ge and
Lin, Chin-Yew and
Zhao, Tiejun",
editor = "Zong, Chengqing and
Xia, Fei and
Li, Wenjie and
Navigli, Roberto",
booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers)",
month = aug,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.acl-short.99/",
doi = "10.18653/v1/2021.acl-short.99",
pages = "786--796",
abstract = "The general format of natural language inference (NLI) makes it tempting to be used for zero-shot text classification by casting any target label into a sentence of hypothesis and verifying whether or not it could be entailed by the input, aiming at generic classification applicable on any specified label space. In this opinion piece, we point out a few overlooked issues that are yet to be discussed in this line of work. We observe huge variance across different classification datasets amongst standard BERT-based NLI models and surprisingly find that pre-trained BERT without any fine-tuning can yield competitive performance against BERT fine-tuned for NLI. With the concern that these models heavily rely on spurious lexical patterns for prediction, we also experiment with preliminary approaches for more robust NLI, but the results are in general negative. Our observations reveal implicit but challenging difficulties in entailment-based zero-shot text classification."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ma-etal-2021-issues">
<titleInfo>
<title>Issues with Entailment-based Zero-shot Text Classification</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tingting</namePart>
<namePart type="family">Ma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jin-Ge</namePart>
<namePart type="family">Yao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chin-Yew</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tiejun</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Chengqing</namePart>
<namePart type="family">Zong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fei</namePart>
<namePart type="family">Xia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wenjie</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Roberto</namePart>
<namePart type="family">Navigli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>The general format of natural language inference (NLI) makes it tempting to be used for zero-shot text classification by casting any target label into a sentence of hypothesis and verifying whether or not it could be entailed by the input, aiming at generic classification applicable on any specified label space. In this opinion piece, we point out a few overlooked issues that are yet to be discussed in this line of work. We observe huge variance across different classification datasets amongst standard BERT-based NLI models and surprisingly find that pre-trained BERT without any fine-tuning can yield competitive performance against BERT fine-tuned for NLI. With the concern that these models heavily rely on spurious lexical patterns for prediction, we also experiment with preliminary approaches for more robust NLI, but the results are in general negative. Our observations reveal implicit but challenging difficulties in entailment-based zero-shot text classification.</abstract>
<identifier type="citekey">ma-etal-2021-issues</identifier>
<identifier type="doi">10.18653/v1/2021.acl-short.99</identifier>
<location>
<url>https://aclanthology.org/2021.acl-short.99/</url>
</location>
<part>
<date>2021-08</date>
<extent unit="page">
<start>786</start>
<end>796</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Issues with Entailment-based Zero-shot Text Classification
%A Ma, Tingting
%A Yao, Jin-Ge
%A Lin, Chin-Yew
%A Zhao, Tiejun
%Y Zong, Chengqing
%Y Xia, Fei
%Y Li, Wenjie
%Y Navigli, Roberto
%S Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers)
%D 2021
%8 August
%I Association for Computational Linguistics
%C Online
%F ma-etal-2021-issues
%X The general format of natural language inference (NLI) makes it tempting to be used for zero-shot text classification by casting any target label into a sentence of hypothesis and verifying whether or not it could be entailed by the input, aiming at generic classification applicable on any specified label space. In this opinion piece, we point out a few overlooked issues that are yet to be discussed in this line of work. We observe huge variance across different classification datasets amongst standard BERT-based NLI models and surprisingly find that pre-trained BERT without any fine-tuning can yield competitive performance against BERT fine-tuned for NLI. With the concern that these models heavily rely on spurious lexical patterns for prediction, we also experiment with preliminary approaches for more robust NLI, but the results are in general negative. Our observations reveal implicit but challenging difficulties in entailment-based zero-shot text classification.
%R 10.18653/v1/2021.acl-short.99
%U https://aclanthology.org/2021.acl-short.99/
%U https://doi.org/10.18653/v1/2021.acl-short.99
%P 786-796
Markdown (Informal)
[Issues with Entailment-based Zero-shot Text Classification](https://aclanthology.org/2021.acl-short.99/) (Ma et al., ACL-IJCNLP 2021)
ACL
- Tingting Ma, Jin-Ge Yao, Chin-Yew Lin, and Tiejun Zhao. 2021. Issues with Entailment-based Zero-shot Text Classification. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers), pages 786–796, Online. Association for Computational Linguistics.