Acta Scientiarum Naturalium Universitatis Pekinensis ›› 2017, Vol. 53 ›› Issue (2): 287-294.DOI: 10.13209/j.0479-8023.2017.039

• Orginal Article • Previous Articles     Next Articles

A Comparative Study on English-Chinese Machine Transliteration

Enting GAO1, Xiangyu DUAN2,()   

  1. 1. School of Electronic and Information Engineering, Suzhou University of Science and Technology, Suzhou 215009
    2. School of Computer Science and Technology, Soochow University, Suzhou 215006
  • Received:2016-07-29 Revised:2016-10-06 Online:2016-11-28 Published:2017-03-20
  • Contact: Xiangyu DUAN


高恩婷1, 段湘煜2,()   

  1. 1. 苏州科技大学电子与信息工程学院, 苏州 215009
    2. 苏州大学计算机科学与技术学院, 苏州 215006
  • 通讯作者: 段湘煜
  • 基金资助:
    国家自然科学基金(61273319, 61373095)资助


With the aim to study the two main methods on machine transliteration: traditional statistical method and the current prevalent deep neural network method, the authors carry out the comparative study on them with two typical systems per method The experiments show that traditional statistical method and deep neural network method perform comparatively regarding evaluation metrics, while manifest difference on individual transliteration result. A system combination method is proposed to balance the strengths of all systems. Experimental results show that system combination significantly improves the transliteration quality over single system.

Key words: machine transliteration, transliteration alignment, statistical method, deep neural network method


针对机器音译的两种主要方法 —— 传统的基于统计的方法和目前流行的基于深度神经网络的方法, 分别使用两种典型系统进行研究。实验结果显示, 基于统计的方法和基于深度神经网络的方法取得的音译质量在评测指标上相当, 但在具体音译结果上各系统间呈现不一致的输出。使用系统融合的方法来实现各系统间的优势互补。实验结果显示, 系统融合的方法显著优于单系统的音译质量。

关键词: 机器音译, 音译对齐, 统计方法, 深度神经网络方法

CLC Number: