|本期目录/Table of Contents|

[1]周张萍,黄荣城,王博立,等.基于增量式自学习策略的多语言翻译模型[J].厦门大学学报(自然科学版),2019,58(02):170-175.[doi:10.6043/j.issn.0438-0479.201811016]
 ZHOU Zhangping,HUANG Rongcheng,WANG Boli,et al.Multilanguage translation model based on incremental self-learning strategy[J].Journal of Xiamen University(Natural Science),2019,58(02):170-175.[doi:10.6043/j.issn.0438-0479.201811016]
点击复制

基于增量式自学习策略的多语言翻译模型(PDF/HTML)
分享到:

《厦门大学学报(自然科学版)》[ISSN:0438-0479/CN:35-1070/N]

卷:
58卷
期数:
2019年02期
页码:
170-175
栏目:
机器翻译模型
出版日期:
2019-03-27

文章信息/Info

Title:
Multilanguage translation model based on incremental self-learning strategy
文章编号:
0438-0479(2019)02-0170-06
作者:
周张萍黄荣城王博立胡金铭史晓东*陈毅东
厦门大学信息科学与技术学院,福建 厦门 360001
Author(s):
ZHOU ZhangpingHUANG RongchengWANG BoliHU JinmingSHI XiaodongCHEN Yidong
School of Information Science and Engineering,Xiamen University,Xiamen 360001,China
关键词:
神经网络机器翻译 多语言机器翻译 增量式自学习
Keywords:
neural machine translation multilingual machine translation iterative method
分类号:
TP 391
DOI:
10.6043/j.issn.0438-0479.201811016
文献标志码:
A
摘要:
针对源语言到目标语言缺乏平行语料的情况,提出了一种基于增量式自学习策略的多语言翻译模型,即利用中介语双语语料训练源语言到目标语言的翻译模型.在Transformer架构下,相比于基于中介语和直接在伪平行语料上训练的普通双语翻译模型,使用该方法在第十四届全国机器翻译研讨会(CWMT 2018)多语言翻译评测数据集上的机器双语互译评估(BLEU)值提升了0.98个百分点.在此基础上,还对比了不同的预处理方法、训练策略以及多模型的平均和集成策略,其中多模型集成策略的BLEU值上可在多模型策略的基础上进一步提升0.53个百分点.
Abstract:
Without parallel corpus from the source language to the target language,we train multilingual neural machine translation models on bilingual corpus of the pivot language and propose an incremental learning strategy to improve source-language to target-language translation.Experimental results under Transformer framework show that our multilingual iterative method can improve the BLEU score by 0.98 percent point on the China workshop on machine translation(CWMT)2018 multi-language translation evaluation data set,compared to traditional pivot-based translation and the vanilla multilingual neural machine translation(NMT).In addition,we also compared different preprocessing methods,training strategies,multi-model average and ensemble,where multi-model ensemble can further increase the BLEU score by 0.53 percent point unpon common multi-model strategy.

参考文献/References:

[1] GOLLINS T,SANDERSON M.Improving cross language retrieval with triangulated translation[C]∥International ACM SIGIR Conference on Research and Development in Information Retrieval.New Orleans:ACM,2001:90-95.
[2] KISHIDA K,KANDO N.Two-stage refinement of query translation in a pivot language approach to cross-lingual information retrieval:an experiment at CLEF 2003[C]∥ Comparative Evaluation of Multilingual Information Access Systems,Workshop of the Cross-Language Evaluation Forum,CLEF 2003.Trondheim:DBLP,2004:253-262.
[3] WU H,WANG H.Pivot language approach for phrase-based statistical machine translation[J].Machine Translation,2007,21(3):165-181.
[4] CHENG Y,YANG Q,LIU Y,et al.Joint training for pivot-based neural machine translation[C]∥ Twenty-Sixth International Joint Conference on Artificial Intelligence.Melbourne:IJCAI,2017:3974-3980.
[5] DONG D,WU H,HE W,et al.Multi-task learning for multiple language translation[C]∥ Meeting of the Association for Computational Linguistics and the International Joint Conference on Natural Language Processing.Beijing:ACL,2015:1723-1732.
[6] LUONG M T,LE Q V,SUTSKEVER I,et al.Multi-task sequence to sequence learning[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1511.06114.
[7] SENNRICH R,HADDOW B,BIRCH A.Improving neural machine translation models with monolingual data[EB/OL].[2016-06-03].http:∥arxiv.org/pdf/1511.06709.
[8] LIU D,ZHU C,ZHAO T,et al.Pivot-based semantic splicing for neural machine translation[C]∥Communications in Computer and Information Science.Singapore:Springer,2016:14-24.doi:10.1007/978-981-10-3635-4_2.
[9] ZOPH B,YURET D,MAY J,et al.Transfer learning for low-resource neural machine translation[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1604.02201.
[10] JOHNSON M,SCHUSTER M,LE Q V,et al.Google’s multilingual neural machine translation system:enabling zero-shot translation[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1611.04558.
[11] HA T L,NIEHUES J,WAIBEL A.Toward multilingual neural machine translation with universal encoder and decoder[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1611.04798.
[12] HA T L,NIEHUES J,WAIBEL A.Effective strategies in zero-shot neural machine translation[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1711.07893.
[13] BECHARA H,MA Y,,GENABITH J V.Statistical post-editing for a statistical MT system[C]∥Proceedings of the Thirteenth Machine Translation Summit(MT Summit XIII).Xiamen:AAMT,2011:308-315.
[14] BERTOLDI N,FEDERICO M.Domain adaptation for statistical machine translation with monolingual resources[C]∥The Workshop on Statistical Machine Translation.Association for Computational Linguistics.Athens:ACL,2009:182-189.
[15] XIA Y,HE D,QIN T,et al.Dual learning for machinetranslation[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1611.00179.
[16] LAKEW S M,LOTITO Q F,NEGRI M,et al.Improving zero-shot translation of low-resource languages[EB/OL].[2018-11-01].http∥arxiv.org/pdf/1811.0138901.
[17] SENNRICH R.,HADDOW B.,BIRCH A.Neural machine translation of rare words with subword units[C]∥ Proceedings of ACL.Berlin:ACL,2016:1715-1725.
[18] PAPINENI K,ROUKOS S,WARD T,et al.IBM research report bleu:a method for automatic evaluation of machine translation[J].Proceedings of Annual Meeting of the Association for Computational Linguistics,2002,30(2):311-318.
[19] LIU Y,ZHOU L,WANG Y,et al.A comparable study on model averaging,ensembling and reranking in NMT[C]∥CCF International Conference on Natural Language Processing and Chinese Computing.Cham:Springer,2018:299-308.[1] GOLLINS T,SANDERSON M.Improving cross language retrieval with triangulated translation[C]∥International ACM SIGIR Conference on Research and Development in Information Retrieval.New Orleans:ACM,2001:90-95.
[2] KISHIDA K,KANDO N.Two-stage refinement of query translation in a pivot language approach to cross-lingual information retrieval:an experiment at CLEF 2003[C]∥ Comparative Evaluation of Multilingual Information Access Systems,Workshop of the Cross-Language Evaluation Forum,CLEF 2003.Trondheim:DBLP,2004:253-262.
[3] WU H,WANG H.Pivot language approach for phrase-based statistical machine translation[J].Machine Translation,2007,21(3):165-181.
[4] CHENG Y,YANG Q,LIU Y,et al.Joint training for pivot-based neural machine translation[C]∥ Twenty-Sixth International Joint Conference on Artificial Intelligence.Melbourne:IJCAI,2017:3974-3980.
[5] DONG D,WU H,HE W,et al.Multi-task learning for multiple language translation[C]∥ Meeting of the Association for Computational Linguistics and the International Joint Conference on Natural Language Processing.Beijing:ACL,2015:1723-1732.
[6] LUONG M T,LE Q V,SUTSKEVER I,et al.Multi-task sequence to sequence learning[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1511.06114.
[7] SENNRICH R,HADDOW B,BIRCH A.Improving neural machine translation models with monolingual data[EB/OL].[2016-06-03].http:∥arxiv.org/pdf/1511.06709.
[8] LIU D,ZHU C,ZHAO T,et al.Pivot-based semantic splicing for neural machine translation[C]∥Communications in Computer and Information Science.Singapore:Springer,2016:14-24.doi:10.1007/978-981-10-3635-4_2.
[9] ZOPH B,YURET D,MAY J,et al.Transfer learning for low-resource neural machine translation[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1604.02201.
[10] JOHNSON M,SCHUSTER M,LE Q V,et al.Google’s multilingual neural machine translation system:enabling zero-shot translation[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1611.04558.
[11] HA T L,NIEHUES J,WAIBEL A.Toward multilingual neural machine translation with universal encoder and decoder[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1611.04798.
[12] HA T L,NIEHUES J,WAIBEL A.Effective strategies in zero-shot neural machine translation[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1711.07893.
[13] BECHARA H,MA Y,,GENABITH J V.Statistical post-editing for a statistical MT system[C]∥Proceedings of the Thirteenth Machine Translation Summit(MT Summit XIII).Xiamen:AAMT,2011:308-315.
[14] BERTOLDI N,FEDERICO M.Domain adaptation for statistical machine translation with monolingual resources[C]∥The Workshop on Statistical Machine Translation.Association for Computational Linguistics.Athens:ACL,2009:182-189.
[15] XIA Y,HE D,QIN T,et al.Dual learning for machinetranslation[EB/OL].[2018-11-01].http:∥arxiv.org/pdf/1611.00179.
[16] LAKEW S M,LOTITO Q F,NEGRI M,et al.Improving zero-shot translation of low-resource languages[EB/OL].[2018-11-01].http∥arxiv.org/pdf/1811.0138901.
[17] SENNRICH R.,HADDOW B.,BIRCH A.Neural machine translation of rare words with subword units[C]∥ Proceedings of ACL.Berlin:ACL,2016:1715-1725.
[18] PAPINENI K,ROUKOS S,WARD T,et al.IBM research report bleu:a method for automatic evaluation of machine translation[J].Proceedings of Annual Meeting of the Association for Computational Linguistics,2002,30(2):311-318.
[19] LIU Y,ZHOU L,WANG Y,et al.A comparable study on model averaging,ensembling and reranking in NMT[C]∥CCF International Conference on Natural Language Processing and Chinese Computing.Cham:Springer,2018:299-308.

备注/Memo

备注/Memo:
收稿日期:2018-11-10 录用日期:2019-02-18
基金项目:国家科技支撑计划项目(2012BAH14F03); 国家自然科学基金(61573294); 教育部博士点基金(20130121110040); 国家语委委托项目(WT135-10); 国家语委甲骨文专项(YWZ-J010)
*通信作者:mandel@xmu.edu.cn
更新日期/Last Update: 1900-01-01