@inproceedings{li-li-2023-enhanced,
title = "Enhanced Training Methods for Multiple Languages",
author = "Li, Hai and
Li, Yang",
editor = "Muresan, Smaranda and
Chen, Vivian and
Casey, Kennington and
David, Vandyke and
Nina, Dethlefs and
Koji, Inoue and
Erik, Ekstedt and
Stefan, Ultes",
booktitle = "Proceedings of the Third DialDoc Workshop on Document-grounded Dialogue and Conversational Question Answering",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.dialdoc-1.6/",
doi = "10.18653/v1/2023.dialdoc-1.6",
pages = "52--56",
abstract = "Document-grounded dialogue generation based on multilingual is a challenging and realistic task. Unlike previous tasks, it need to tackle with multiple high-resource languages facilitating low-resource languages. This paper summarizes our research based on a three-stage pipeline that includes retrieval, re-rank and generation where each component is individually optimized. In different languages with limited data scenarios, we mainly improve the robustness of the pipeline through data augmentation and embedding perturbation with purpose of improving the performance designing three training methods: cross-language enhancement training, weighted training with neighborhood distribution augmentation, and ensemble adversarial training, all of that can be used as plug and play modules. Through experiments with different settings, it has been shown that our methods can effectively improve the generalization performance of pipeline with score ranking 6th among the public submissions on leaderboards."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-li-2023-enhanced">
<titleInfo>
<title>Enhanced Training Methods for Multiple Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hai</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Third DialDoc Workshop on Document-grounded Dialogue and Conversational Question Answering</title>
</titleInfo>
<name type="personal">
<namePart type="given">Smaranda</namePart>
<namePart type="family">Muresan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivian</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kennington</namePart>
<namePart type="family">Casey</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vandyke</namePart>
<namePart type="family">David</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dethlefs</namePart>
<namePart type="family">Nina</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Inoue</namePart>
<namePart type="family">Koji</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekstedt</namePart>
<namePart type="family">Erik</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ultes</namePart>
<namePart type="family">Stefan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Toronto, Canada</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Document-grounded dialogue generation based on multilingual is a challenging and realistic task. Unlike previous tasks, it need to tackle with multiple high-resource languages facilitating low-resource languages. This paper summarizes our research based on a three-stage pipeline that includes retrieval, re-rank and generation where each component is individually optimized. In different languages with limited data scenarios, we mainly improve the robustness of the pipeline through data augmentation and embedding perturbation with purpose of improving the performance designing three training methods: cross-language enhancement training, weighted training with neighborhood distribution augmentation, and ensemble adversarial training, all of that can be used as plug and play modules. Through experiments with different settings, it has been shown that our methods can effectively improve the generalization performance of pipeline with score ranking 6th among the public submissions on leaderboards.</abstract>
<identifier type="citekey">li-li-2023-enhanced</identifier>
<identifier type="doi">10.18653/v1/2023.dialdoc-1.6</identifier>
<location>
<url>https://aclanthology.org/2023.dialdoc-1.6/</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>52</start>
<end>56</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Enhanced Training Methods for Multiple Languages
%A Li, Hai
%A Li, Yang
%Y Muresan, Smaranda
%Y Chen, Vivian
%Y Casey, Kennington
%Y David, Vandyke
%Y Nina, Dethlefs
%Y Koji, Inoue
%Y Erik, Ekstedt
%Y Stefan, Ultes
%S Proceedings of the Third DialDoc Workshop on Document-grounded Dialogue and Conversational Question Answering
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F li-li-2023-enhanced
%X Document-grounded dialogue generation based on multilingual is a challenging and realistic task. Unlike previous tasks, it need to tackle with multiple high-resource languages facilitating low-resource languages. This paper summarizes our research based on a three-stage pipeline that includes retrieval, re-rank and generation where each component is individually optimized. In different languages with limited data scenarios, we mainly improve the robustness of the pipeline through data augmentation and embedding perturbation with purpose of improving the performance designing three training methods: cross-language enhancement training, weighted training with neighborhood distribution augmentation, and ensemble adversarial training, all of that can be used as plug and play modules. Through experiments with different settings, it has been shown that our methods can effectively improve the generalization performance of pipeline with score ranking 6th among the public submissions on leaderboards.
%R 10.18653/v1/2023.dialdoc-1.6
%U https://aclanthology.org/2023.dialdoc-1.6/
%U https://doi.org/10.18653/v1/2023.dialdoc-1.6
%P 52-56
Markdown (Informal)
[Enhanced Training Methods for Multiple Languages](https://aclanthology.org/2023.dialdoc-1.6/) (Li & Li, dialdoc 2023)
ACL
- Hai Li and Yang Li. 2023. Enhanced Training Methods for Multiple Languages. In Proceedings of the Third DialDoc Workshop on Document-grounded Dialogue and Conversational Question Answering, pages 52–56, Toronto, Canada. Association for Computational Linguistics.