典型文献
Document-Level Neural Machine Translation with Hierarchical Modeling of Global Context
文献摘要：
Document-level machine translation(MT)remains challenging due to its difficulty in efficiently using document-level global context for translation.In this paper,we propose a hierarchical model to learn the global context for document-level neural machine translation(NMT).This is done through a sentence encoder to capture intra-sentence dependencies and a document encoder to model document-level inter-sentence consistency and coherence.With this hierarchical architec-ture,we feedback the extracted document-level global context to each word in a top-down fashion to distinguish different translations of a word according to its specific surrounding context.Notably,we explore the effect of three popular atten-tion functions during the information backward-distribution phase to take a deep look into the global context information distribution of our model.In addition,since large-scale in-domain document-level parallel corpora are usually unavailable,we use a two-step training strategy to take advantage of a large-scale corpus with out-of-domain parallel sentence pairs and a small-scale corpus with in-domain parallel document pairs to achieve the domain adaptability.Experimental results of our model on Chinese-English and English-German corpora significantly improve the Transformer baseline by 4.5 BLEU points on average which demonstrates the effectiveness of our proposed hierarchical model in document-level NMT.
文献关键词：
中图分类号：
作者姓名：
Xin Tan;Long-Yin Zhang;Guo-Dong Zhou
作者机构：
School of Computer Science and Technology,Soochow University,Suzhou 215006,China
文献出处：
引用格式：
[1]Xin Tan;Long-Yin Zhang;Guo-Dong Zhou-.Document-Level Neural Machine Translation with Hierarchical Modeling of Global Context)[J].计算机科学技术学报（英文版）,2022(02):295-308
A类：
B类：
Document,Level,Neural,Machine,Translation,Hierarchical,Modeling,Global,Context,level,machine,remains,challenging,due,its,difficulty,efficiently,using,document,global,context,In,this,paper,we,hierarchical,model,learn,neural,NMT,This,done,through,sentence,encoder,capture,intra,dependencies,inter,consistency,coherence,With,architec,feedback,extracted,each,word,top,down,fashion,distinguish,different,translations,according,specific,surrounding,Notably,explore,three,popular,atten,functions,during,information,backward,distribution,phase,take,deep,look,into,our,addition,since,large,scale,domain,parallel,corpora,are,usually,unavailable,use,two,step,training,strategy,advantage,corpus,out,pairs,small,achieve,adaptability,Experimental,results,Chinese,English,German,significantly,improve,Transformer,baseline,by,BLEU,points,average,which,demonstrates,effectiveness,proposed
AB值：
0.568895
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
