Research on Mongolian-Chinese Machine Translation Annotated with Gated Recurrent Unit Part of Speech
With the development and progress of Information Technology, the translation between different languages has become particularly important. Statistical Machine Translation may be able to predict a relatively accurate target word with statistical analysis method, but it cannot get a much better translation as it couldn’t fully understand the semantic context information of source language words. In order to solve this problem, the model of Mongolian-Chinese Machine Translation System could be constructed based on GRU (Gated Recurrent Unit) neural network structure and the usage of global attention mechanism to obtain bilingual alignment information. In the process of constructing a dictionary, the bilingual words are annotated to further improve the alignment probability. The research result shows that the BLEU value is certainly promoted and improved compared with previous benchmark research and traditional statistical machine translation method.
KeywordsMachine translation Gated Recurrent Unit (GRU) Attention mechanism Alignment bilingual Part of speech
This work was partially funded by the National Natural Science Foundation of China (61363052, 61502255), Natural Science Foundation of Inner Mongolia Autonomous Region funded projects (2016MS0605), Inner Mongolia Autonomous Region Ethnic Affairs Committee Fund funded projects (MW-2017-MGYWXXH-03).
- 1.Du, Z.P., Wu, B.M., Zhang, L.H., et al.: Rule-based POS tagging in English-Chinese MT system. J. Inf. Eng. Univ. 4(3), 89–92 (2003)Google Scholar
- 2.Na, B.Q.: A Statistical Machine Translation System Between Mongolian and Chinese. Inner Mongolia University, Hohhot (2006)Google Scholar
- 3.Yang, N.: Neural Network Learning for Statistical Machine Translation. University of Science and Technology of China, Hefei (2014)Google Scholar
- 4.Zhen, T.Q.: Research of the Mixed Language Model Based on Statistical and Latent Semantic Analysis. Harbin Institute of Technology, Harbin (2007)Google Scholar
- 5.Cho, K., Merrienboer, B.V., Bahdanau, D., et al.: On the properties of neural machine translation: encoder-decoder approaches. Comput. Sci. 2(12), 55–68 (2014)Google Scholar
- 6.Jean, S., Cho, K., Memisevic, R., et al.: On using very large target vocabulary for neural machine translation. Comput. Sci. 7(4), 11–23 (2014)Google Scholar
- 7.Sundermeyer M, Schlüter R, Ney H. LSTM Neural Networks for Language Modeling[C]// Interspeech. 2012:601–608Google Scholar
- 9.Wang, S.R.G.L., Si, Q.T., Na, S.W.R.T.: Phrase-based Chinese Mongolian statistical machine translation. Comput. Eng. Appl. 46(14), 138–142 (2010)Google Scholar
- 10.Wang, R.L., Luan, J., Pan, X.H., et al.: An improved forward maximum matching algorithm for Chinese word segmentation. Comput. Appl. Softw. 28(3), 195–197 (2011)Google Scholar
- 11.Li, R., Zheng, J.H.: Application Study of improved Viterbi algorithm. Comput. Eng. Des. 28(3), 530–531 (2007)Google Scholar
- 12.Wiseman, S., Chopra, S., Ranzato, M., et al.: Training language models using target-propagation 2(11), 79–92 (2017)Google Scholar