@inproceedings{kurosawa-etal-2018-japanese,
title = "{J}apanese Predicate Conjugation for Neural Machine Translation",
author = "Kurosawa, Michiki and
Matsumura, Yukio and
Yamagishi, Hayahide and
Komachi, Mamoru",
editor = "Cordeiro, Silvio Ricardo and
Oraby, Shereen and
Pavalanathan, Umashanthi and
Rim, Kyeongmin",
booktitle = "Proceedings of the 2018 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Student Research Workshop",
month = jun,
year = "2018",
address = "New Orleans, Louisiana, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/N18-4014",
doi = "10.18653/v1/N18-4014",
pages = "100--105",
abstract = "Neural machine translation (NMT) has a drawback in that can generate only high-frequency words owing to the computational costs of the softmax function in the output layer. In Japanese-English NMT, Japanese predicate conjugation causes an increase in vocabulary size. For example, one verb can have as many as 19 surface varieties. In this research, we focus on predicate conjugation for compressing the vocabulary size in Japanese. The vocabulary list is filled with the various forms of verbs. We propose methods using predicate conjugation information without discarding linguistic information. The proposed methods can generate low-frequency words and deal with unknown words. Two methods were considered to introduce conjugation information: the first considers it as a token (conjugation token) and the second considers it as an embedded vector (conjugation feature). The results using these methods demonstrate that the vocabulary size can be compressed by approximately 86.1{\%} (Tanaka corpus) and the NMT models can output the words not in the training data set. Furthermore, BLEU scores improved by 0.91 points in Japanese-to-English translation, and 0.32 points in English-to-Japanese translation with ASPEC.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kurosawa-etal-2018-japanese">
<titleInfo>
<title>Japanese Predicate Conjugation for Neural Machine Translation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Michiki</namePart>
<namePart type="family">Kurosawa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yukio</namePart>
<namePart type="family">Matsumura</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hayahide</namePart>
<namePart type="family">Yamagishi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mamoru</namePart>
<namePart type="family">Komachi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2018-06</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop</title>
</titleInfo>
<name type="personal">
<namePart type="given">Silvio</namePart>
<namePart type="given">Ricardo</namePart>
<namePart type="family">Cordeiro</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shereen</namePart>
<namePart type="family">Oraby</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Umashanthi</namePart>
<namePart type="family">Pavalanathan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kyeongmin</namePart>
<namePart type="family">Rim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">New Orleans, Louisiana, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Neural machine translation (NMT) has a drawback in that can generate only high-frequency words owing to the computational costs of the softmax function in the output layer. In Japanese-English NMT, Japanese predicate conjugation causes an increase in vocabulary size. For example, one verb can have as many as 19 surface varieties. In this research, we focus on predicate conjugation for compressing the vocabulary size in Japanese. The vocabulary list is filled with the various forms of verbs. We propose methods using predicate conjugation information without discarding linguistic information. The proposed methods can generate low-frequency words and deal with unknown words. Two methods were considered to introduce conjugation information: the first considers it as a token (conjugation token) and the second considers it as an embedded vector (conjugation feature). The results using these methods demonstrate that the vocabulary size can be compressed by approximately 86.1% (Tanaka corpus) and the NMT models can output the words not in the training data set. Furthermore, BLEU scores improved by 0.91 points in Japanese-to-English translation, and 0.32 points in English-to-Japanese translation with ASPEC.</abstract>
<identifier type="citekey">kurosawa-etal-2018-japanese</identifier>
<identifier type="doi">10.18653/v1/N18-4014</identifier>
<location>
<url>https://aclanthology.org/N18-4014</url>
</location>
<part>
<date>2018-06</date>
<extent unit="page">
<start>100</start>
<end>105</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Japanese Predicate Conjugation for Neural Machine Translation
%A Kurosawa, Michiki
%A Matsumura, Yukio
%A Yamagishi, Hayahide
%A Komachi, Mamoru
%Y Cordeiro, Silvio Ricardo
%Y Oraby, Shereen
%Y Pavalanathan, Umashanthi
%Y Rim, Kyeongmin
%S Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop
%D 2018
%8 June
%I Association for Computational Linguistics
%C New Orleans, Louisiana, USA
%F kurosawa-etal-2018-japanese
%X Neural machine translation (NMT) has a drawback in that can generate only high-frequency words owing to the computational costs of the softmax function in the output layer. In Japanese-English NMT, Japanese predicate conjugation causes an increase in vocabulary size. For example, one verb can have as many as 19 surface varieties. In this research, we focus on predicate conjugation for compressing the vocabulary size in Japanese. The vocabulary list is filled with the various forms of verbs. We propose methods using predicate conjugation information without discarding linguistic information. The proposed methods can generate low-frequency words and deal with unknown words. Two methods were considered to introduce conjugation information: the first considers it as a token (conjugation token) and the second considers it as an embedded vector (conjugation feature). The results using these methods demonstrate that the vocabulary size can be compressed by approximately 86.1% (Tanaka corpus) and the NMT models can output the words not in the training data set. Furthermore, BLEU scores improved by 0.91 points in Japanese-to-English translation, and 0.32 points in English-to-Japanese translation with ASPEC.
%R 10.18653/v1/N18-4014
%U https://aclanthology.org/N18-4014
%U https://doi.org/10.18653/v1/N18-4014
%P 100-105
Markdown (Informal)
[Japanese Predicate Conjugation for Neural Machine Translation](https://aclanthology.org/N18-4014) (Kurosawa et al., NAACL 2018)
ACL
- Michiki Kurosawa, Yukio Matsumura, Hayahide Yamagishi, and Mamoru Komachi. 2018. Japanese Predicate Conjugation for Neural Machine Translation. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop, pages 100–105, New Orleans, Louisiana, USA. Association for Computational Linguistics.