| アイテムタイプ |
会議発表論文 / Conference Paper(1) |
| 公開日 |
2025-07-24 |
| タイトル |
|
|
タイトル |
Centroid-Based Efficient Minimum Bayes Risk Decoding |
| 言語 |
|
|
言語 |
eng |
| 資源タイプ |
|
|
資源タイプ |
conference paper |
| アクセス権 |
|
|
アクセス権 |
open access |
| 著者 |
Deguchi, Hiroyuki
坂井, 優介
上垣外, 英剛
渡辺, 太郎
Tanaka, Hideki
Utiyama, Masao
|
| 抄録 |
|
|
内容記述タイプ |
Abstract |
|
内容記述 |
Minimum Bayes risk (MBR) decoding achieved state-of-the-art translation performance by using COMET, a neural metric that has a high correlation with human evaluation.However, MBR decoding requires quadratic time since it computes the expected score between a translation hypothesis and all reference translations.We propose centroid-based MBR (CBMBR) decoding to improve the speed of MBR decoding.Our method clusters the reference translations in the feature space, and then calculates the score using the centroids of each cluster.The experimental results show that our CBMBR not only improved the decoding speed of the expected score calculation 5.7 times, but also outperformed vanilla MBR decoding in translation quality by up to 0.5 COMET in the WMT’22 En$2194Ja, En$2194De, En$2194Zh, and WMT’23 En$2194Ja translation tasks. |
| 書誌情報 |
en : Findings of the Association for Computational Linguistics ACL 2024
p. 11009-11018,
ページ数 10,
発行日 2024-08
|
| 会議情報 |
|
|
|
会議名 |
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) |
|
回次 |
62 |
|
|
開始年 |
2024 |
|
|
開始月 |
08 |
|
|
開始日 |
11 |
|
|
終了年 |
2024 |
|
|
終了月 |
08 |
|
|
終了日 |
16 |
|
|
開催期間 |
2024-08-11 - 2024-08-16 |
|
|
開催地 |
Bangkok, Thailand |
|
開催国 |
THA |
| 出版者 |
|
|
出版者 |
Association for Computational Linguistics |
| 出版者版DOI |
|
|
関連タイプ |
isReplacedBy |
|
|
識別子タイプ |
DOI |
|
|
関連識別子 |
https://doi.org/10.18653/v1/2024.findings-acl.654 |
| 出版者版URI |
|
|
関連タイプ |
isReplacedBy |
|
|
識別子タイプ |
URI |
|
|
関連識別子 |
https://aclanthology.org/2024.findings-acl.654/ |
| 権利 |
|
|
権利情報Resource |
https://creativecommons.org/licenses/by/4.0/ |
|
権利情報 |
$00A92024 Association for Computational Linguistics. ACL materials are Copyright $00A9 1963$20132025 ACL; other materials are copyrighted by their respective copyright holders. Materials prior to 2016 here are licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 International License. Permission is granted to make copies for the purposes of teaching and research. Materials published in or after 2016 are licensed on a Creative Commons Attribution 4.0 International License. |
| 著者版フラグ |
|
|
出版タイプ |
NA |
| 助成情報 |
|
|
|
助成機関名 |
Japan Society for the Promotion of Science (JSPS) |
|
|
研究課題番号 |
JP21H05054 |
|
|
研究課題名 |
多元自動通訳システムと評価法に関する研究とその応用展開 |
