计算机工程与应用 ›› 2012, Vol. 48 ›› Issue (4): 29-32.

• 博士论坛 • 上一篇    下一篇

一种混合策略的专利机器翻译系统研究

晋耀红1,2   

  1. 1.北京师范大学 中文信息处理研究所,北京 100875
    2.中国专利信息中心-北京师范大学机器翻译联合实验室,北京 100875
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2012-02-01 发布日期:2012-04-05

Hybrid-strategy method combining semantic analysis with rule-based MT for patent machine translation

JIN Yaohong1,2   

  1. 1.Institute of Chinese Information Processing, Beijing Normal University, Beijing 100875, China
    2.CPIC-BNU Joint Laboratory for Machine Translation, Beijing 100875, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2012-02-01 Published:2012-04-05

摘要: 针对专利文本翻译中的复杂语句,提出了一种基于混合策略的方法,融合语义分析技术和基于规则的翻译技术,来提高专利翻译的效果。利用语义分析技术,重点解决句子中心动词识别和句子中有嵌套结构存在的名称短语的分析,把语义分析结果输入到基于规则的翻译系统中,用以改善翻译的效果。测试结果表明,融合后的翻译系统,BLEU值提高了9.8%。该方法已经集成到了国家知识产权局的在线汉英机器翻译系统中,有效地提高了专利翻译的效果和翻译效率。

关键词: 语义分析, 机器翻译, 语义特征, 概念层次网络, 专利文献

Abstract: Focusing on the complex sentence translation in patent machine translation, this paper presents a hybrid method combining semantic analysis with rule-based MT. The semantic analysis is used to deal with the main verb recognition and the boundary detection of long NP. The result of semantic analysis is input to the rule-based MT system to improve the performance of translation. The result of the experiments shows that this hybrid-strategy method can effectively improve the BLEU score 9.8%. The method has been integrated into an online patent MT system running in SIPO.

Key words: semantic analysis, machine translation, semantic feature, Hierarchical Network of Concepts(HNC), patent