Extracting translation equivalences automatically based on tree-string

被引：0

作者：

Zhang C.-X.

Liang Y.-H. ^{[1
]}

Li P. ^{[2
]}

Lu Z.-M. ^{[3
]}

Liu Y. ^{[4
]}

机构：

[1] School of Computer Engineering, Vocational University of Suzhou City

[2] College of Computer Science and Technology, Harbin University of Science and Technology

[3] College of Information and Communication Engineering, Harbin Engineering University

[4] School of Computer Science and Technology, Harbin Institute of Technology

来源：

Information Technology Journal | 2010年 / 9卷 / 02期

关键词：

Parsing information; Translation equivalence; Word alignment result;

D O I：

10.3923/itj.2010.371.375

中图分类号：

学科分类号：

摘要：

In this study, we propose a new method to align Chinese-English bilingual sentence pairs based on parsing tree of Chinese sentence and word alignment result. From tree-string alignment, translation equivalences are extracted. Experimental results indicate that the new method achieves 80.97% at precision, which is better than parse-parse-match method. The performance of tree-string alignment is better than that of parse-parse-match on extracting translation equivalences. © 2010 Asian Network for Scientific Information.

引用

页码：371 / 375

页数：4