@inproceedings{gu-feng-2020-investigating,
title = "Investigating Catastrophic Forgetting During Continual Training for Neural Machine Translation",
author = "Gu, Shuhao and
Feng, Yang",
booktitle = "Proceedings of the 28th International Conference on Computational Linguistics",
month = dec,
year = "2020",
address = "Barcelona, Spain (Online)",
publisher = "International Committee on Computational Linguistics",
url = "https://aclanthology.org/2020.coling-main.381",
doi = "10.18653/v1/2020.coling-main.381",
pages = "4315--4326",
abstract = "Neural machine translation (NMT) models usually suffer from catastrophic forgetting during continual training where the models tend to gradually forget previously learned knowledge and swing to fit the newly added data which may have a different distribution, e.g. a different domain. Although many methods have been proposed to solve this problem, we cannot get to know what causes this phenomenon yet. Under the background of domain adaptation, we investigate the cause of catastrophic forgetting from the perspectives of modules and parameters (neurons). The investigation on the modules of the NMT model shows that some modules have tight relation with the general-domain knowledge while some other modules are more essential in the domain adaptation. And the investigation on the parameters shows that some parameters are important for both the general-domain and in-domain translation and the great change of them during continual training brings about the performance decline in general-domain. We conducted experiments across different language pairs and domains to ensure the validity and reliability of our findings.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="gu-feng-2020-investigating">
<titleInfo>
<title>Investigating Catastrophic Forgetting During Continual Training for Neural Machine Translation</title>
</titleInfo>
<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">Yang</namePart>
<namePart type="family">Feng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2020-dec</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 28th International Conference on Computational Linguistics</title>
</titleInfo>
<originInfo>
<publisher>International Committee on Computational Linguistics</publisher>
<place>
<placeTerm type="text">Barcelona, Spain (Online)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Neural machine translation (NMT) models usually suffer from catastrophic forgetting during continual training where the models tend to gradually forget previously learned knowledge and swing to fit the newly added data which may have a different distribution, e.g. a different domain. Although many methods have been proposed to solve this problem, we cannot get to know what causes this phenomenon yet. Under the background of domain adaptation, we investigate the cause of catastrophic forgetting from the perspectives of modules and parameters (neurons). The investigation on the modules of the NMT model shows that some modules have tight relation with the general-domain knowledge while some other modules are more essential in the domain adaptation. And the investigation on the parameters shows that some parameters are important for both the general-domain and in-domain translation and the great change of them during continual training brings about the performance decline in general-domain. We conducted experiments across different language pairs and domains to ensure the validity and reliability of our findings.</abstract>
<identifier type="citekey">gu-feng-2020-investigating</identifier>
<identifier type="doi">10.18653/v1/2020.coling-main.381</identifier>
<location>
<url>https://aclanthology.org/2020.coling-main.381</url>
</location>
<part>
<date>2020-dec</date>
<extent unit="page">
<start>4315</start>
<end>4326</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Investigating Catastrophic Forgetting During Continual Training for Neural Machine Translation
%A Gu, Shuhao
%A Feng, Yang
%S Proceedings of the 28th International Conference on Computational Linguistics
%D 2020
%8 dec
%I International Committee on Computational Linguistics
%C Barcelona, Spain (Online)
%F gu-feng-2020-investigating
%X Neural machine translation (NMT) models usually suffer from catastrophic forgetting during continual training where the models tend to gradually forget previously learned knowledge and swing to fit the newly added data which may have a different distribution, e.g. a different domain. Although many methods have been proposed to solve this problem, we cannot get to know what causes this phenomenon yet. Under the background of domain adaptation, we investigate the cause of catastrophic forgetting from the perspectives of modules and parameters (neurons). The investigation on the modules of the NMT model shows that some modules have tight relation with the general-domain knowledge while some other modules are more essential in the domain adaptation. And the investigation on the parameters shows that some parameters are important for both the general-domain and in-domain translation and the great change of them during continual training brings about the performance decline in general-domain. We conducted experiments across different language pairs and domains to ensure the validity and reliability of our findings.
%R 10.18653/v1/2020.coling-main.381
%U https://aclanthology.org/2020.coling-main.381
%U https://doi.org/10.18653/v1/2020.coling-main.381
%P 4315-4326
Markdown (Informal)
[Investigating Catastrophic Forgetting During Continual Training for Neural Machine Translation](https://aclanthology.org/2020.coling-main.381) (Gu & Feng, COLING 2020)
ACL