On the Properties of Neural Machine Translation: Encoder-Decoder
Approaches
K. Cho, B. van Merrienboer, D. Bahdanau, and Y. Bengio. (2014)cite arxiv:1409.1259Comment: Eighth Workshop on Syntax, Semantics and Structure in Statistical Translation (SSST-8).
Abstract
Neural machine translation is a relatively new approach to statistical
machine translation based purely on neural networks. The neural machine
translation models often consist of an encoder and a decoder. The encoder
extracts a fixed-length representation from a variable-length input sentence,
and the decoder generates a correct translation from this representation. In
this paper, we focus on analyzing the properties of the neural machine
translation using two models; RNN Encoder--Decoder and a newly proposed gated
recursive convolutional neural network. We show that the neural machine
translation performs relatively well on short sentences without unknown words,
but its performance degrades rapidly as the length of the sentence and the
number of unknown words increase. Furthermore, we find that the proposed gated
recursive convolutional network learns a grammatical structure of a sentence
automatically.
%0 Generic
%1 cho2014properties
%A Cho, Kyunghyun
%A van Merrienboer, Bart
%A Bahdanau, Dzmitry
%A Bengio, Yoshua
%D 2014
%K GRU sem_wise2223
%T On the Properties of Neural Machine Translation: Encoder-Decoder
Approaches
%U http://arxiv.org/abs/1409.1259
%X Neural machine translation is a relatively new approach to statistical
machine translation based purely on neural networks. The neural machine
translation models often consist of an encoder and a decoder. The encoder
extracts a fixed-length representation from a variable-length input sentence,
and the decoder generates a correct translation from this representation. In
this paper, we focus on analyzing the properties of the neural machine
translation using two models; RNN Encoder--Decoder and a newly proposed gated
recursive convolutional neural network. We show that the neural machine
translation performs relatively well on short sentences without unknown words,
but its performance degrades rapidly as the length of the sentence and the
number of unknown words increase. Furthermore, we find that the proposed gated
recursive convolutional network learns a grammatical structure of a sentence
automatically.
@misc{cho2014properties,
abstract = {Neural machine translation is a relatively new approach to statistical
machine translation based purely on neural networks. The neural machine
translation models often consist of an encoder and a decoder. The encoder
extracts a fixed-length representation from a variable-length input sentence,
and the decoder generates a correct translation from this representation. In
this paper, we focus on analyzing the properties of the neural machine
translation using two models; RNN Encoder--Decoder and a newly proposed gated
recursive convolutional neural network. We show that the neural machine
translation performs relatively well on short sentences without unknown words,
but its performance degrades rapidly as the length of the sentence and the
number of unknown words increase. Furthermore, we find that the proposed gated
recursive convolutional network learns a grammatical structure of a sentence
automatically.},
added-at = {2022-09-13T14:01:41.000+0200},
author = {Cho, Kyunghyun and van Merrienboer, Bart and Bahdanau, Dzmitry and Bengio, Yoshua},
biburl = {https://www.bibsonomy.org/bibtex/272efe495d5909fee3b0e5a2266955fa1/annakrause},
description = {1409.1259.pdf},
interhash = {762b86f271d539b0c572e58174401369},
intrahash = {72efe495d5909fee3b0e5a2266955fa1},
keywords = {GRU sem_wise2223},
note = {cite arxiv:1409.1259Comment: Eighth Workshop on Syntax, Semantics and Structure in Statistical Translation (SSST-8)},
timestamp = {2022-09-13T14:01:41.000+0200},
title = {On the Properties of Neural Machine Translation: Encoder-Decoder
Approaches},
url = {http://arxiv.org/abs/1409.1259},
year = 2014
}