摘要
针对大量英文地名急需实现中文翻译及目前英语地名自动翻译研究较少、翻译结果质量低下的问题,该文提出了一种英语地名机器翻译方法。首先对于不同类别地名语料基于点互信息和有向无环图数据结构提取地名模板,然后基于统计模板模型解析地名词法结构实现翻译。实验结果表明:该方法翻译精度相对较高,可有效应用于地名生产。本方法解决了地名翻译中通专名合理区分的难题,同时利用地名类别属性辅助翻译,大大节省了人工成本,对我国地名翻译及生产领域具有一定参考意义。
In order to solve the problem that a large number of English place names need to be translated into Chinese urgently and that there are few studies on automatic translation of English place names at present,and the quality of translation results is low,a method of machine translation of English place names is proposed in this paper.Firstly,the Toponymic template is extracted from different types of Toponymic corpus based on point mutual information and directed acyclic graph data structure,and then the Toponymic structure is parsed and translated based on statistical template model.The experimental results show that the method has relatively high translation accuracy and can be effectively used in toponym production.This method solves the problem of distinguishing proper names in place name translation.At the same time,it uses category attributes of place names to assist translation,which greatly saves the labor cost and has a certain reference significance in the field of place name translation and production in China.
作者
任洪凯
毛曦
王继周
马维军
殷红梅
REN Hongkai;MAO Xi;WANG Jizhou;MA Weijun;YIN Hongmei(Southern Marine Science and Engineering Guangdong Laboratory(Zhuhai),Zhuhai,Guangdong 519082;Chinese Academy of Surveying&Mapping,Beijing 100036,China;Capital Normal University,Beijing 100048,China)
出处
《测绘科学》
CSCD
北大核心
2021年第2期152-158,共7页
Science of Surveying and Mapping
基金
自然资源部重大项目(12113600000018)
科技部重点研发项目(2017-2021)(2017YFB0503600)
中国测绘科学研究院基本科研业务经费项目(AR1912)。
关键词
地名
机器翻译
类别
点互信息
有向无环图
词法结构解析
place name
machine translation
category
point mutual information
directed acyclic graph
lexical structure analysis
