摘要
针对目前海图英语地名机器翻译研究薄弱、传统人工翻译效率低等问题,提出了一种语料库和规则驱动相结合的海图英语地名机器翻译方法:首先分析海图英语地名组成特点及翻译难点,在此基础上设计英汉海图地名双语语料库及通名表,实现对现有海图地名资料的继承和地名通名的自动识别与翻译;而后结合英语地名译音表及译写细则开展音节切分,进而采用正向最大匹配算法完成待译字符串与英汉音译表的匹配,实现英语地名专名音译;最后组合输出地名通名及专名译写结果。经过对海图英语地名翻译实验,验证了该机器翻译方法的有效性,为大规模海图英语地名翻译奠定基础。
Aiming at weakness of machine translation of English Place Names for Charts and the efficiency of traditional manual translation is low,this paper proposed a machine translation scheme of English Place Names for Charts based on corpus and rule-driven mechnism:firstly,by analyzing the compositional characteristics and translation difficulties of English Place Names for Charts,the transliteration table and translation rules of English Place Names for Charts are constructed,in order to realize the inheritance of available data of place names and automatic recognition and translation of place generic names.Secondly,based on syllable division and forward maximum matching algorithm,English proper place names automatic transliteration was realized by mechanical matching between the splitted character strings and entries of the English and Chinese transliteration table.Finally,the result of translation of generic and proper place names are combined and output.Through the experiments of English Place Names for Charts translation,the validity of the machine translation method has been verified,which makes it possible to produce large-scale place names.
作者
乔晶
李鹤元
QIAO Jing;LI Heyuan(Studying Abroad Programme in XiJ an International Studies University,Xi*an 710061,China;Xi'an Surveying and Mapping Center,Xi'an 710054,China)
出处
《海洋测绘》
CSCD
北大核心
2022年第5期73-77,82,共6页
Hydrographic Surveying and Charting
基金
国家自然科学基金(42171456)。
关键词
海图英语地名
语料库
翻译规则库
音节切分
最大正向匹配算法
English place names for charts
corpus
translation rules
syllable division
forward maximum matching
