基于DOP技术实现英汉机译  

IMPLEMENTING ENGLISH-CHINESE MACHINE TRANSLATION BASED ON DOP TECHNIQUE

作　　者：张玥杰[1] 朱靖波[2] 姚天顺[2] 

机构地区：[1]复旦大学计算机科学与工程系,上海200433 [2]东北大学计算机科学与工程系,沈阳110006

出　　处：《模式识别与人工智能》2003年第2期129-135,共7页Pattern Recognition and Artificial Intelligence

基　　金：国家自然科学基金(No.60203010);国家863计划(No.2001AA114120)资助项目

摘　　要：本文以DOP理论作为语言现象的描述基础,以英汉机译为背景,对面向数据的英汉机译方法进行深入研究。其中,不仅提出一种基于DOP技术的源语组合分析模式,而且建立了相对应的目标语生成机制。前者以DOP技术作为基本框架,通过源语语句片段组合形式的获取以及基于相似的概率评估,完成整个分析过程。而后者基于传统上主要应用于分析过程的DOP技术,通过对分析结果进行线性化操作,最终生成目标语译文。实验表明,目标语译文质量比较令人满意,可成功实现英汉机译。Taking the DOP theory as the description basis of linguistics phenomena and English-Chinese machine translation as the research ground, we make a thorough and painstaking research work about data-oriented English-Chinese machine translation method. In this paper, a kind of source language combination analysis model based on DOP technique is proposed. And a kind of target language generation mechanism correspondent with the source language combination analysis model is also established. The former takes DOP technique as the basic framework. Through the acquisition of all the fragment-combination-forms of the source language sentence and the similarity-based probability estimation, the whole source language combination analysis procedure is completed. The latter applies DOP technique which is used in language analysis traditionally into target language generation equally. Through linearizing source language analysis result-syntax tree, the final translation in target language is generated. The experiment result shows that the quality of translation in target language is satisfactory and the English-Chinese machine translation process can be implemented successfully.

关 键 词：机器翻译 自然语言处理 DOP 语言学 英语 汉语 计算机 源语组合分析 

分 类 号：TP391.2[自动化与计算机技术—计算机应用技术]