@inproceedings{wang-etal-2022-improving,
title = "Improving Faithfulness by Augmenting Negative Summaries from Fake Documents",
author = "Wang, Tianshu and
Ladhak, Faisal and
Durmus, Esin and
He, He",
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.816",
doi = "10.18653/v1/2022.emnlp-main.816",
pages = "11913--11921",
abstract = "Current abstractive summarization systems tend to hallucinate content that is unfaithful to the source document, posing a risk of misinformation. To mitigate hallucination, we must teach the model to distinguish hallucinated summaries from faithful ones. However, the commonly used maximum likelihood training does not disentangle factual errors from other model errors. To address this issue,we propose a back-translation-style approach to augment negative samples that mimic factual errors made by the model. Specifically, we train an elaboration model that generates hallucinated documents given the reference summaries, and then generates negative summaries from the fake documents. We incorporate the negative samples into training through a controlled generator, which produces faithful/unfaithful summaries conditioned on the control codes. Additionally, we find that adding textual entailment data through multitasking further boosts the performance. Experiments on three datasets (XSum, Gigaword, and WikiHow) show that our method consistently improves faithfulness without sacrificing informativeness according to both human and automatic evaluation",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2022-improving">
<titleInfo>
<title>Improving Faithfulness by Augmenting Negative Summaries from Fake Documents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tianshu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Faisal</namePart>
<namePart type="family">Ladhak</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Esin</namePart>
<namePart type="family">Durmus</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">He</namePart>
<namePart type="family">He</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>Current abstractive summarization systems tend to hallucinate content that is unfaithful to the source document, posing a risk of misinformation. To mitigate hallucination, we must teach the model to distinguish hallucinated summaries from faithful ones. However, the commonly used maximum likelihood training does not disentangle factual errors from other model errors. To address this issue,we propose a back-translation-style approach to augment negative samples that mimic factual errors made by the model. Specifically, we train an elaboration model that generates hallucinated documents given the reference summaries, and then generates negative summaries from the fake documents. We incorporate the negative samples into training through a controlled generator, which produces faithful/unfaithful summaries conditioned on the control codes. Additionally, we find that adding textual entailment data through multitasking further boosts the performance. Experiments on three datasets (XSum, Gigaword, and WikiHow) show that our method consistently improves faithfulness without sacrificing informativeness according to both human and automatic evaluation</abstract>
<identifier type="citekey">wang-etal-2022-improving</identifier>
<identifier type="doi">10.18653/v1/2022.emnlp-main.816</identifier>
<location>
<url>https://aclanthology.org/2022.emnlp-main.816</url>
</location>
<part>
<date>2022-12</date>
<extent unit="page">
<start>11913</start>
<end>11921</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Improving Faithfulness by Augmenting Negative Summaries from Fake Documents
%A Wang, Tianshu
%A Ladhak, Faisal
%A Durmus, Esin
%A He, He
%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 wang-etal-2022-improving
%X Current abstractive summarization systems tend to hallucinate content that is unfaithful to the source document, posing a risk of misinformation. To mitigate hallucination, we must teach the model to distinguish hallucinated summaries from faithful ones. However, the commonly used maximum likelihood training does not disentangle factual errors from other model errors. To address this issue,we propose a back-translation-style approach to augment negative samples that mimic factual errors made by the model. Specifically, we train an elaboration model that generates hallucinated documents given the reference summaries, and then generates negative summaries from the fake documents. We incorporate the negative samples into training through a controlled generator, which produces faithful/unfaithful summaries conditioned on the control codes. Additionally, we find that adding textual entailment data through multitasking further boosts the performance. Experiments on three datasets (XSum, Gigaword, and WikiHow) show that our method consistently improves faithfulness without sacrificing informativeness according to both human and automatic evaluation
%R 10.18653/v1/2022.emnlp-main.816
%U https://aclanthology.org/2022.emnlp-main.816
%U https://doi.org/10.18653/v1/2022.emnlp-main.816
%P 11913-11921
Markdown (Informal)
[Improving Faithfulness by Augmenting Negative Summaries from Fake Documents](https://aclanthology.org/2022.emnlp-main.816) (Wang et al., EMNLP 2022)
ACL