期刊
NEUROCOMPUTING
卷 284, 期 -, 页码 138-147出版社
ELSEVIER
DOI: 10.1016/j.neucom.2018.01.010
关键词
Neural Machine Translation; Word lattice; Recurrent Neural Network; Gated Recurrent Unit
资金
- Natural Science Foundation of China [61573294, 61672440]
- Ph.D. Programs Foundation of Ministry of Education of China [20130121110040]
- Foundation of the State Language Commission of China [WT135-10, YB135-49]
- Natural Science Foundation of Fujian Province [2016J05161]
- Fund of Research Project of Tibet Autonomous Region of China [Z2014A18G2-13]
- National Key Technology RD Program [2012BAH14F03]
The dominant Neural Machine Translation (NMT) models usually resort to word-level modeling to embed input sentences into semantic space. However, it may not be optimal for the encoder modeling of NMT, especially for languages where tokenizations are usually ambiguous: On one hand, there may be tokenization errors which may negatively affect the encoder modeling of NMT. On the other hand, the optimal tokenization granularity is unclear for NMT. In this paper, we propose lattice-to-sequence attentional NMT models, which generalize the standard Recurrent Neural Network (RNN) encoders to lattice topology. Specifically, they take as input a word lattice which compactly encodes many tokenization alternatives, and learn to generate the hidden state for the current step from multiple inputs and hidden states in previous steps. Compared with the standard RNN encoder, the proposed encoders not only alleviate the negative impact of tokenization errors but are more expressive and flexible as well for encoding the meaning of input sentences. Experimental results on both Chinese-English and Japanese-English translations demonstrate the effectiveness of our models. (C) 2018 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据