@inproceedings{zhuocheng-etal-2023-addressing,
title = "Addressing the Length Bias Challenge in Document-Level Neural Machine Translation",
author = "Zhuocheng, Zhang and
Gu, Shuhao and
Zhang, Min and
Feng, Yang",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2023",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.findings-emnlp.773/",
doi = "10.18653/v1/2023.findings-emnlp.773",
pages = "11545--11556",
abstract = "Document-level neural machine translation (DNMT) has shown promising results by incorporating context information through increased maximum lengths of source and target sentences. However, this approach also introduces a length bias problem, whereby DNMT suffers from significant translation quality degradation when decoding sentences that are much shorter or longer than the maximum sentence length during training, i.e., the length bias problem. To prevent the model from neglecting shorter sentences, we sample the training data to ensure a more uniform distribution across different sentence lengths while progressively increasing the maximum sentence length during training. Additionally, we introduce a length-normalized attention mechanism to aid the model in focusing on target information, mitigating the issue of attention divergence when processing longer sentences. Furthermore, during the decoding stage of DNMT, we propose a sliding decoding strategy that limits the length of target sentences to not exceed the maximum length encountered during training. The experimental results indicate that our method can achieve state-of-the-art results on several open datasets, and further analysis shows that our method can significantly alleviate the length bias problem."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhuocheng-etal-2023-addressing">
<titleInfo>
<title>Addressing the Length Bias Challenge in Document-Level Neural Machine Translation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zhang</namePart>
<namePart type="family">Zhuocheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shuhao</namePart>
<namePart type="family">Gu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Min</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Feng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2023</title>
</titleInfo>
<name type="personal">
<namePart type="given">Houda</namePart>
<namePart type="family">Bouamor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Juan</namePart>
<namePart type="family">Pino</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kalika</namePart>
<namePart type="family">Bali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Singapore</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Document-level neural machine translation (DNMT) has shown promising results by incorporating context information through increased maximum lengths of source and target sentences. However, this approach also introduces a length bias problem, whereby DNMT suffers from significant translation quality degradation when decoding sentences that are much shorter or longer than the maximum sentence length during training, i.e., the length bias problem. To prevent the model from neglecting shorter sentences, we sample the training data to ensure a more uniform distribution across different sentence lengths while progressively increasing the maximum sentence length during training. Additionally, we introduce a length-normalized attention mechanism to aid the model in focusing on target information, mitigating the issue of attention divergence when processing longer sentences. Furthermore, during the decoding stage of DNMT, we propose a sliding decoding strategy that limits the length of target sentences to not exceed the maximum length encountered during training. The experimental results indicate that our method can achieve state-of-the-art results on several open datasets, and further analysis shows that our method can significantly alleviate the length bias problem.</abstract>
<identifier type="citekey">zhuocheng-etal-2023-addressing</identifier>
<identifier type="doi">10.18653/v1/2023.findings-emnlp.773</identifier>
<location>
<url>https://aclanthology.org/2023.findings-emnlp.773/</url>
</location>
<part>
<date>2023-12</date>
<extent unit="page">
<start>11545</start>
<end>11556</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Addressing the Length Bias Challenge in Document-Level Neural Machine Translation
%A Zhuocheng, Zhang
%A Gu, Shuhao
%A Zhang, Min
%A Feng, Yang
%Y Bouamor, Houda
%Y Pino, Juan
%Y Bali, Kalika
%S Findings of the Association for Computational Linguistics: EMNLP 2023
%D 2023
%8 December
%I Association for Computational Linguistics
%C Singapore
%F zhuocheng-etal-2023-addressing
%X Document-level neural machine translation (DNMT) has shown promising results by incorporating context information through increased maximum lengths of source and target sentences. However, this approach also introduces a length bias problem, whereby DNMT suffers from significant translation quality degradation when decoding sentences that are much shorter or longer than the maximum sentence length during training, i.e., the length bias problem. To prevent the model from neglecting shorter sentences, we sample the training data to ensure a more uniform distribution across different sentence lengths while progressively increasing the maximum sentence length during training. Additionally, we introduce a length-normalized attention mechanism to aid the model in focusing on target information, mitigating the issue of attention divergence when processing longer sentences. Furthermore, during the decoding stage of DNMT, we propose a sliding decoding strategy that limits the length of target sentences to not exceed the maximum length encountered during training. The experimental results indicate that our method can achieve state-of-the-art results on several open datasets, and further analysis shows that our method can significantly alleviate the length bias problem.
%R 10.18653/v1/2023.findings-emnlp.773
%U https://aclanthology.org/2023.findings-emnlp.773/
%U https://doi.org/10.18653/v1/2023.findings-emnlp.773
%P 11545-11556
Markdown (Informal)
[Addressing the Length Bias Challenge in Document-Level Neural Machine Translation](https://aclanthology.org/2023.findings-emnlp.773/) (Zhuocheng et al., Findings 2023)
ACL