A Teacher-Student Framework for Zero-Resource Neural Machine Translation

被引：75

作者：

Chen, Yun ^{[1
]}

Liu, Yang ^{[2
,3
]}

Cheng, Yong ^{[4
]}

Li, Victor O. K. ^{[1
]}

机构：

[1] Univ Hong Kong, Dept Elect & Elect Engn, Hong Kong, Peoples R China

[2] Tsinghua Univ, Dept Comp Sci & Technol, Tsinghua Natl Lab Informat Sci & Technol, State Key Lab Intelligent Technol & Syst, Beijing, Peoples R China

[3] Jiangsu Collaborat Innovat Ctr Language Competenc, Xuzhou, Jiangsu, Peoples R China

[4] Tsinghua Univ, Inst Interdisciplinary Informat Sci, Beijing, Peoples R China

来源：

PROCEEDINGS OF THE 55TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2017), VOL 1 | 2017年

基金：

中国国家自然科学基金;

关键词：

D O I：

10.18653/v1/P17-1176

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

While end-to-end neural machine translation (NMT) has made remarkable progress recently, it still suffers from the data scarcity problem for low-resource language pairs and domains. In this paper, we propose a method for zero-resource NMT by assuming that parallel sentences have close probabilities of generating a sentence in a third language. Based on the assumption, our method is able to train a source-to-target NMT model ("student") without parallel corpora available guided by an existing pivot-to-target NMT model ("teacher") on a source-pivot parallel corpus. Experimental results show that the proposed method significantly improves over a baseline pivot-based model by +3.0 BLEU points across various language pairs.

引用

页码：1925 / 1935

页数：11