面向英汉机器翻译的英语特征语义块处理  

Processing of English eigen chunk for English-Chinese machine translation

作　　者：董凌冲[1] 李颖[1] 池毓焕[2] 

机构地区：[1]装甲兵工程学院信息工程系,北京100072 [2]中国科学院声学研究所,北京100080

出　　处：《计算机工程与设计》2010年第14期3260-3263,3267,共5页Computer Engineering and Design

基　　金：国家科技支撑计划基金项目(2007BAH05B02-05)

摘　　要：针对当前英汉机器翻译软件未能充分理解源语句的问题,基于概念层次网络理论(hierarchical network of concepts,HNC理论),详细分析了在语句理解中起重要作用的英语特征语义块,分类讨论了其核心部分构成特点,并提出了相应的计算机处理算法。与在线机器翻译结果的比较表明,该算法对源语句进行了更加深入的分析与理解,能够为英汉机器翻译提供技术支持。To address the source language understanding problem of current English-Chinese machine translation software,eigen chunk （EK） structure which is of great importance in English sentence understanding is studied. Based on HNC theory,characteristics of English EK structure,especially the kernel part are analyzed. And the processing algorithm is put forward. By making comparison with online machine translation,the results show that the algorithm proposed is able to conduct in-depth analysis to get better understanding of the source sentence. So it can provide technical support for English-Chinese machine translation.

关 键 词：特征语义块构成 概念层次网络理论 英汉机器翻译 自然语言处理 句类分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]