@inproceedings{song-etal-2021-alignart,
title = "{A}lig{NART}: Non-autoregressive Neural Machine Translation by Jointly Learning to Estimate Alignment and Translate",
author = "Song, Jongyoon and
Kim, Sungwon and
Yoon, Sungroh",
editor = "Moens, Marie-Francine and
Huang, Xuanjing and
Specia, Lucia and
Yih, Scott Wen-tau",
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2021",
address = "Online and Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.emnlp-main.1/",
doi = "10.18653/v1/2021.emnlp-main.1",
pages = "1--14",
abstract = "Non-autoregressive neural machine translation (NART) models suffer from the multi-modality problem which causes translation inconsistency such as token repetition. Most recent approaches have attempted to solve this problem by implicitly modeling dependencies between outputs. In this paper, we introduce AligNART, which leverages full alignment information to explicitly reduce the modality of the target distribution. AligNART divides the machine translation task into (i) alignment estimation and (ii) translation with aligned decoder inputs, guiding the decoder to focus on simplified one-to-one translation. To alleviate the alignment estimation problem, we further propose a novel alignment decomposition method. Our experiments show that AligNART outperforms previous non-iterative NART models that focus on explicit modality reduction on WMT14 En{\ensuremath{\leftrightarrow}}De and WMT16 Ro{\textrightarrow}En. Furthermore, AligNART achieves BLEU scores comparable to those of the state-of-the-art connectionist temporal classification based models on WMT14 En{\ensuremath{\leftrightarrow}}De. We also observe that AligNART effectively addresses the token repetition problem even without sequence-level knowledge distillation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="song-etal-2021-alignart">
<titleInfo>
<title>AligNART: Non-autoregressive Neural Machine Translation by Jointly Learning to Estimate Alignment and Translate</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jongyoon</namePart>
<namePart type="family">Song</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sungwon</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sungroh</namePart>
<namePart type="family">Yoon</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Marie-Francine</namePart>
<namePart type="family">Moens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuanjing</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lucia</namePart>
<namePart type="family">Specia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Scott</namePart>
<namePart type="given">Wen-tau</namePart>
<namePart type="family">Yih</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online and Punta Cana, Dominican Republic</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Non-autoregressive neural machine translation (NART) models suffer from the multi-modality problem which causes translation inconsistency such as token repetition. Most recent approaches have attempted to solve this problem by implicitly modeling dependencies between outputs. In this paper, we introduce AligNART, which leverages full alignment information to explicitly reduce the modality of the target distribution. AligNART divides the machine translation task into (i) alignment estimation and (ii) translation with aligned decoder inputs, guiding the decoder to focus on simplified one-to-one translation. To alleviate the alignment estimation problem, we further propose a novel alignment decomposition method. Our experiments show that AligNART outperforms previous non-iterative NART models that focus on explicit modality reduction on WMT14 En\ensuremathłeftrightarrowDe and WMT16 Ro→En. Furthermore, AligNART achieves BLEU scores comparable to those of the state-of-the-art connectionist temporal classification based models on WMT14 En\ensuremathłeftrightarrowDe. We also observe that AligNART effectively addresses the token repetition problem even without sequence-level knowledge distillation.</abstract>
<identifier type="citekey">song-etal-2021-alignart</identifier>
<identifier type="doi">10.18653/v1/2021.emnlp-main.1</identifier>
<location>
<url>https://aclanthology.org/2021.emnlp-main.1/</url>
</location>
<part>
<date>2021-11</date>
<extent unit="page">
<start>1</start>
<end>14</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T AligNART: Non-autoregressive Neural Machine Translation by Jointly Learning to Estimate Alignment and Translate
%A Song, Jongyoon
%A Kim, Sungwon
%A Yoon, Sungroh
%Y Moens, Marie-Francine
%Y Huang, Xuanjing
%Y Specia, Lucia
%Y Yih, Scott Wen-tau
%S Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
%D 2021
%8 November
%I Association for Computational Linguistics
%C Online and Punta Cana, Dominican Republic
%F song-etal-2021-alignart
%X Non-autoregressive neural machine translation (NART) models suffer from the multi-modality problem which causes translation inconsistency such as token repetition. Most recent approaches have attempted to solve this problem by implicitly modeling dependencies between outputs. In this paper, we introduce AligNART, which leverages full alignment information to explicitly reduce the modality of the target distribution. AligNART divides the machine translation task into (i) alignment estimation and (ii) translation with aligned decoder inputs, guiding the decoder to focus on simplified one-to-one translation. To alleviate the alignment estimation problem, we further propose a novel alignment decomposition method. Our experiments show that AligNART outperforms previous non-iterative NART models that focus on explicit modality reduction on WMT14 En\ensuremathłeftrightarrowDe and WMT16 Ro→En. Furthermore, AligNART achieves BLEU scores comparable to those of the state-of-the-art connectionist temporal classification based models on WMT14 En\ensuremathłeftrightarrowDe. We also observe that AligNART effectively addresses the token repetition problem even without sequence-level knowledge distillation.
%R 10.18653/v1/2021.emnlp-main.1
%U https://aclanthology.org/2021.emnlp-main.1/
%U https://doi.org/10.18653/v1/2021.emnlp-main.1
%P 1-14
Markdown (Informal)
[AligNART: Non-autoregressive Neural Machine Translation by Jointly Learning to Estimate Alignment and Translate](https://aclanthology.org/2021.emnlp-main.1/) (Song et al., EMNLP 2021)
ACL