Machine Translation Based on Translation Rules for Processing Natural Language

Abstract:

This paper presents a hybrid approach, which integrates an example-pattern-based method and a rule-based method, to the design and implementation of an English-Chinese machine translation system. It focuses discussion on language model , knowledge base, design ideas and implementation strategies. Our system has been tested based on requirement details listed in the Outlines for Automatic Evaluation of Machine Translation constituted by National Hi-Tech Project 863, and compared with the Huajian system. Experiment results indicate that our system has high translation speed and accuracy.

• Periodical: Advanced Materials Research (Volumes 926-930)
• Pages: 2682-2685
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.926-930.2682
• Online since: May 2014
• Authors: Lan Dong Li*, Wen Ying Xing, Xue Long Zhang
• Keywords: Example-Pattern-Based, Hybrid, Machine Translation System, Natural Language Processing, Rule-Based

References

[1] B.M. Wu: Matlink English - Chinese machine translation system (Ph.D., Zhengzhou : Information Engineering University of PLA, China 2010), p.34.

[2] T.J. Zhao: Principle of machine translation (Harbin Institute of Technology Press, China 2009).

[3] Y.H. Guo: Computer applications, Vol. 24(2008) No. 3, p.71.

[4] Z.P. Du: Journal of Information Engineering University, Vol. 4(2010) No. 3, p.89.