英汉机器翻译中基于规则的句子结构分析与转换  

Rule-Based Sentence Structure Analysis and Transformation for E-C MT

作　　者：吴保民[1] 郭永辉[2] 王炳锡[3] 

机构地区：[1]信息工程大学理学院,河南郑州450001 [2]信息工程大学电子技术学院,河南郑州450004 [3]信息工程大学信息工程学院,河南郑州450002

出　　处：《信息工程大学学报》2007年第1期30-33,共4页Journal of Information Engineering University

基　　金：国家自然科学基金资助项目(60372038)

摘　　要：以英汉机器翻译系统为研究背景,提出了一种基于规则的句子结构分析与转换方法。针对不同情况采取不同策略将复杂句分解成由一组简单句构成的复杂句子结构。简单句结构分析以谓语动词为中心,依据其EICG格约束和句子浅层分析结果填充格框架,从而得到整个句子的语法语义结构。句子结构转换依照句子分析结果和转换规则产生等价的目标语言的线性语法语义结构。系统运行结果表明,本文提出的方法能够正确有效地分析英语句子的结构并完成英汉句子的转换。This paper is intended to present a rule-based syntactical analysis and transformation ap- proach for English-Chinese machine translation system. Long and complex sentences are first decomposed into simple-sentence structures for further analysis. Analysis of predicate verb in a simple sentence is carried out based on the EICG constraints and shallow syntactic parsing results to get the syntactic and semantic＇ structure of the whole sentence. The equivalent syntactic and semantic transformation into the target language is based on analytical results and transformation rules. The system has been tested against the Outlines for Automatic Evaluation of Machine Translation constituted by National Hi-Tech Project 863, and compared with Huajian system. Experimental results demonstrate high translation accuracy and quick translation speed.

关 键 词：机器翻译 句子结构分析 结构转换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]