首页 | 本学科首页   官方微博 | 高级检索  
     检索      

面向专利领域的汉英机器翻译融合系统
引用本文:李洪政,赵凯,胡韧奋,蒋宏飞,朱筠,晋耀红.面向专利领域的汉英机器翻译融合系统[J].情报工程,2017,3(3):105-115.
作者姓名:李洪政  赵凯  胡韧奋  蒋宏飞  朱筠  晋耀红
作者单位:北京师范大学中文信息处理研究所,北京奇虎科技有限公司,北京师范大学中文信息处理研究所,中科鼎富(北京)科技发展有限公司,北京师范大学中文信息处理研究所,中科鼎富(北京)科技发展有限公司
基金项目:本文受国家高技术研究发展计划基金项目“海量文本多层次知识表示及中文文本理解应用系统研制”(2012AA011104)的资助。
摘    要:面向专利领域的机器翻译近年来已成为机器翻译的重要应用领域之一。本文提出了一个汉英专利文本机器翻译融合系统,该系统以规则系统为主导搭建,并把规则翻译方法和基于短语的统计翻译系统相结合。在融合系统中,规则系统主要负责源语言的分析和转换阶段的处理,生成相应的源语言句法分析树与转换树,并确定目标语言的基本句法框架。统计翻译系统则在目标语生成阶段根据生成的目标语句法结构寻找合适的对译词形,并产生最终的候选译文。通过利用自动评测指标对融合系统进行测试,融合系统的结果均优于单个规则系统和统计系统的结果,表明了融合方法的有效性和可行性,可以改善系统的翻译性能,提高翻译质量。

关 键 词:专利,规则方法,统计方法,融合系统,机器翻译

A Hybrid System for Chinese-English Patent Machine Translation
Authors:LI HongZheng  ZHAO Kai  HU RenFen  JIANG HongFei  ZHU Yun and JIN YaoHong
Institution:Institute of Chinese Information Processing, Beijing Normal University,Beijing Qihu Technology Co., Ltd.,,Institute of Chinese Information Processing, Beijing Normal University,Beijing Dinfo Technology Co., Ltd.,,Institute of Chinese Information Processing, Beijing Normal University and Beijing Dinfo Technology Co., Ltd.,
Abstract:Machine translation towards patent domain has become one important application of machine translation in recent years. This paper presented a novel hybrid system, which combines rule-based machine translation (RBMT) with phrase-based statistical machine translation (SMT), to translate Chinese patent texts into English. The hybrid architecture is basically guided by the RBMT engine which processes source language parsing and transformation, generating proper syntactic trees for the target language. In the generation stage, the SMT subsystem provides proper lexical selection according to the generated grammatical structures and produces final translation. According to the experimental evaluation, the hybrid approach outperforms each individual system over sets of automatic evaluation metrics, indicating that the proposed method performs well in improving translation results.
Keywords:Patent  rule-based  statistical-based  hybrid system  machine translation
点击此处可从《情报工程》浏览原始摘要信息
点击此处可从《情报工程》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号