Advertisement

Research on Machine Translation Model Based on Neural Network

  • Zhuoran HanEmail author
  • Shenghong Li
Conference paper
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 517)

Abstract

Machine Translation is an important part of Natural Language Processing. The model based on convolution neural network and attention mechanism (Fcnn model), which was proposed by Facebook in 2017, has been successful. We use this Fcnn model as the baseline model of this subject, and on the base of this model, we try to improve it by the combination of bytes pair encoding method, model ensemble method. In this issue, we use Bilingual Evaluation Understudy (BLEU) as a criterion to measure the quality of translation. After testing, these methods can improve the translation quality of the model. Finally, the overall translation quality increased from 0.28 of the baseline Fcnn model to 0.32.

Keywords

Machine translation CNN BLEU 

References

  1. 1.
    Krizhevsky, A., Sutskever, I., Hinton, G.E.: ImageNet classification with deep convolutional neural networks. Commun. ACM 60(2), 2012 (2012)Google Scholar
  2. 2.
    Och, F.J., Ney, H.: Discriminative training and maximum entropy models for statistical machine translation. In: Meeting on Association for Computational Linguistics. Association for Computational Linguistics, pp. 295–302 (2002)Google Scholar
  3. 3.
    Papineni, K., Roukos, S., Ward, T., et al.: BLEU: a method for automatic evaluation of machine translation. In: Meeting on Association for Computational Linguistics. Association for Computational Linguistics, pp. 311–318 (2002)Google Scholar
  4. 4.
    Mikolov, T., Chen, K., Corrado, G., et al.: Efficient estimation of word representations in vector space. Comput. Sci. (2013)Google Scholar
  5. 5.
    Sutskever, I., Vinyals, O., Le, Q.V.: Sequence to sequence learning with neural networks, vol. 4, pp. 3104–3112 (2014)Google Scholar
  6. 6.
    Bahdanau, D., Cho, K., Bengio, Y.: Neural machine translation by jointly learning to align and translate. Comput. Sci. (2014)Google Scholar
  7. 7.
    Graves, A.: Long short-term memory. Supervised Sequence Labelling with Recurrent Neural Networks, pp. 1735–1780. Springer, Berlin (2012)CrossRefGoogle Scholar
  8. 8.
    Vaswani, A., Bengio, S., Brevdo, E., et al.: Tensor2Tensor for neural machine translation. CoRR (2018)Google Scholar
  9. 9.
    Luong, M.T., Brevdo, E., Zhao, R.: Neural machine translation (seq2seq) tutorial (2017). https://github.com/tensorflow/nmt

Copyright information

© Springer Nature Singapore Pte Ltd. 2020

Authors and Affiliations

  1. 1.Shanghai Jiao Tong UniversityShanghaiChina

Personalised recommendations