@inproceedings{balachandran-etal-2022-correcting,
title = "Correcting Diverse Factual Errors in Abstractive Summarization via Post-Editing and Language Model Infilling",
author = "Balachandran, Vidhisha and
Hajishirzi, Hannaneh and
Cohen, William and
Tsvetkov, Yulia",
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.667",
doi = "10.18653/v1/2022.emnlp-main.667",
pages = "9818--9830",
abstract = "Abstractive summarization models often generate inconsistent summaries containing factual errors or hallucinated content. Recent works focus on correcting factual errors in generated summaries via post-editing. Such correction models are trained using adversarial non-factual summaries constructed using heuristic rules for injecting errors. However, generating non-factual summaries using heuristics often does not generalize well to actual model errors. In this work, we propose to generate hard, representative synthetic examples of non-factual summaries through infilling language models. With this data, we train a more robust fact-correction model to post-edit the summaries to improve factual consistency. Through quantitative and qualitative experiments on two popular summarization datasets{---} CNN/DM and XSum{---}we show that our approach vastly outperforms prior methods in correcting erroneous summaries. Our model{---}FactEdit{---}improves factuality scores by over {\textasciitilde}11 points on CNN/DM and over {\textasciitilde}31 points on XSum on average across multiple summarization models, producing more factual summaries while maintaining competitive summarization quality.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="balachandran-etal-2022-correcting">
<titleInfo>
<title>Correcting Diverse Factual Errors in Abstractive Summarization via Post-Editing and Language Model Infilling</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vidhisha</namePart>
<namePart type="family">Balachandran</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hannaneh</namePart>
<namePart type="family">Hajishirzi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">William</namePart>
<namePart type="family">Cohen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yulia</namePart>
<namePart type="family">Tsvetkov</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>Abstractive summarization models often generate inconsistent summaries containing factual errors or hallucinated content. Recent works focus on correcting factual errors in generated summaries via post-editing. Such correction models are trained using adversarial non-factual summaries constructed using heuristic rules for injecting errors. However, generating non-factual summaries using heuristics often does not generalize well to actual model errors. In this work, we propose to generate hard, representative synthetic examples of non-factual summaries through infilling language models. With this data, we train a more robust fact-correction model to post-edit the summaries to improve factual consistency. Through quantitative and qualitative experiments on two popular summarization datasets— CNN/DM and XSum—we show that our approach vastly outperforms prior methods in correcting erroneous summaries. Our model—FactEdit—improves factuality scores by over ~11 points on CNN/DM and over ~31 points on XSum on average across multiple summarization models, producing more factual summaries while maintaining competitive summarization quality.</abstract>
<identifier type="citekey">balachandran-etal-2022-correcting</identifier>
<identifier type="doi">10.18653/v1/2022.emnlp-main.667</identifier>
<location>
<url>https://aclanthology.org/2022.emnlp-main.667</url>
</location>
<part>
<date>2022-12</date>
<extent unit="page">
<start>9818</start>
<end>9830</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Correcting Diverse Factual Errors in Abstractive Summarization via Post-Editing and Language Model Infilling
%A Balachandran, Vidhisha
%A Hajishirzi, Hannaneh
%A Cohen, William
%A Tsvetkov, Yulia
%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 balachandran-etal-2022-correcting
%X Abstractive summarization models often generate inconsistent summaries containing factual errors or hallucinated content. Recent works focus on correcting factual errors in generated summaries via post-editing. Such correction models are trained using adversarial non-factual summaries constructed using heuristic rules for injecting errors. However, generating non-factual summaries using heuristics often does not generalize well to actual model errors. In this work, we propose to generate hard, representative synthetic examples of non-factual summaries through infilling language models. With this data, we train a more robust fact-correction model to post-edit the summaries to improve factual consistency. Through quantitative and qualitative experiments on two popular summarization datasets— CNN/DM and XSum—we show that our approach vastly outperforms prior methods in correcting erroneous summaries. Our model—FactEdit—improves factuality scores by over ~11 points on CNN/DM and over ~31 points on XSum on average across multiple summarization models, producing more factual summaries while maintaining competitive summarization quality.
%R 10.18653/v1/2022.emnlp-main.667
%U https://aclanthology.org/2022.emnlp-main.667
%U https://doi.org/10.18653/v1/2022.emnlp-main.667
%P 9818-9830
Markdown (Informal)
[Correcting Diverse Factual Errors in Abstractive Summarization via Post-Editing and Language Model Infilling](https://aclanthology.org/2022.emnlp-main.667) (Balachandran et al., EMNLP 2022)
ACL