@inproceedings{udomcharoenchaikit-etal-2022-mitigating,
title = "Mitigating Spurious Correlation in Natural Language Understanding with Counterfactual Inference",
author = "Udomcharoenchaikit, Can and
Ponwitayarat, Wuttikorn and
Payoungkhamdee, Patomporn and
Masuk, Kanruethai and
Buaphet, Weerayut and
Chuangsuwanich, Ekapol and
Nutanong, Sarana",
editor = "Goldberg, Yoav and
Kozareva, Zornitsa and
Zhang, Yue",
booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2022",
address = "Abu Dhabi, United Arab Emirates",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.emnlp-main.777/",
doi = "10.18653/v1/2022.emnlp-main.777",
pages = "11308--11321",
abstract = "Despite their promising results on standard benchmarks, NLU models are still prone to make predictions based on shortcuts caused by unintended bias in the dataset. For example, an NLI model may use lexical overlap as a shortcut to make entailment predictions due to repetitive data generation patterns from annotators, also called annotation artifacts. In this paper, we propose a causal analysis framework to help debias NLU models. We show that (1) by defining causal relationships, we can introspect how much annotation artifacts affect the outcomes. (2) We can utilize counterfactual inference to mitigate bias with this knowledge. We found that viewing a model as a treatment can mitigate bias more effectively than viewing annotation artifacts as treatment. (3) In addition to bias mitigation, we can interpret how much each debiasing strategy is affected by annotation artifacts. Our experimental results show that using counterfactual inference can improve out-of-distribution performance in all settings while maintaining high in-distribution performance."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="udomcharoenchaikit-etal-2022-mitigating">
<titleInfo>
<title>Mitigating Spurious Correlation in Natural Language Understanding with Counterfactual Inference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Can</namePart>
<namePart type="family">Udomcharoenchaikit</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wuttikorn</namePart>
<namePart type="family">Ponwitayarat</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Patomporn</namePart>
<namePart type="family">Payoungkhamdee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kanruethai</namePart>
<namePart type="family">Masuk</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weerayut</namePart>
<namePart type="family">Buaphet</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekapol</namePart>
<namePart type="family">Chuangsuwanich</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sarana</namePart>
<namePart type="family">Nutanong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yoav</namePart>
<namePart type="family">Goldberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zornitsa</namePart>
<namePart type="family">Kozareva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Abu Dhabi, United Arab Emirates</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Despite their promising results on standard benchmarks, NLU models are still prone to make predictions based on shortcuts caused by unintended bias in the dataset. For example, an NLI model may use lexical overlap as a shortcut to make entailment predictions due to repetitive data generation patterns from annotators, also called annotation artifacts. In this paper, we propose a causal analysis framework to help debias NLU models. We show that (1) by defining causal relationships, we can introspect how much annotation artifacts affect the outcomes. (2) We can utilize counterfactual inference to mitigate bias with this knowledge. We found that viewing a model as a treatment can mitigate bias more effectively than viewing annotation artifacts as treatment. (3) In addition to bias mitigation, we can interpret how much each debiasing strategy is affected by annotation artifacts. Our experimental results show that using counterfactual inference can improve out-of-distribution performance in all settings while maintaining high in-distribution performance.</abstract>
<identifier type="citekey">udomcharoenchaikit-etal-2022-mitigating</identifier>
<identifier type="doi">10.18653/v1/2022.emnlp-main.777</identifier>
<location>
<url>https://aclanthology.org/2022.emnlp-main.777/</url>
</location>
<part>
<date>2022-12</date>
<extent unit="page">
<start>11308</start>
<end>11321</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Mitigating Spurious Correlation in Natural Language Understanding with Counterfactual Inference
%A Udomcharoenchaikit, Can
%A Ponwitayarat, Wuttikorn
%A Payoungkhamdee, Patomporn
%A Masuk, Kanruethai
%A Buaphet, Weerayut
%A Chuangsuwanich, Ekapol
%A Nutanong, Sarana
%Y Goldberg, Yoav
%Y Kozareva, Zornitsa
%Y Zhang, Yue
%S Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
%D 2022
%8 December
%I Association for Computational Linguistics
%C Abu Dhabi, United Arab Emirates
%F udomcharoenchaikit-etal-2022-mitigating
%X Despite their promising results on standard benchmarks, NLU models are still prone to make predictions based on shortcuts caused by unintended bias in the dataset. For example, an NLI model may use lexical overlap as a shortcut to make entailment predictions due to repetitive data generation patterns from annotators, also called annotation artifacts. In this paper, we propose a causal analysis framework to help debias NLU models. We show that (1) by defining causal relationships, we can introspect how much annotation artifacts affect the outcomes. (2) We can utilize counterfactual inference to mitigate bias with this knowledge. We found that viewing a model as a treatment can mitigate bias more effectively than viewing annotation artifacts as treatment. (3) In addition to bias mitigation, we can interpret how much each debiasing strategy is affected by annotation artifacts. Our experimental results show that using counterfactual inference can improve out-of-distribution performance in all settings while maintaining high in-distribution performance.
%R 10.18653/v1/2022.emnlp-main.777
%U https://aclanthology.org/2022.emnlp-main.777/
%U https://doi.org/10.18653/v1/2022.emnlp-main.777
%P 11308-11321
Markdown (Informal)
[Mitigating Spurious Correlation in Natural Language Understanding with Counterfactual Inference](https://aclanthology.org/2022.emnlp-main.777/) (Udomcharoenchaikit et al., EMNLP 2022)
ACL