摘要
针对跨语言信息检索中关联英文翻译的选择准确度不高的问题,提出一种基于最关联语义本体模型匹配的跨语言信息检索英文翻译选取方法。首先构建跨语言信息检索中最关联英文语义的本体结构模型,采用语义指向性信息索引方法进行英文翻译的上下文语义映射;然后根据语义本体之间的词语知识和本体片段映射方法进行英文语义翻译的特征提取,实现最关联英文语义翻译选取;最后进行实验测试分析。结果表明,采用该方法进行跨语言信息检索,英文语义翻译选取的召回性能较好,查全率、查准率较高。
Aiming at the problem that the selection accuracy of relevant English translation in the cross?language information retrieval is not high,a selection method of English translation for cross?language information retrieval based on most relevant semantic ontology model matching is proposed.The most relevant English semantic ontology structure model for cross?language information retrieval is built first,then the semantic directivity information index method is used to make context semantic mapping of English translation,and the feature extraction of English semantic translation is conducted according to word&expression knowledge and ontology fragment mapping method to realize the relevant English semantic translation selection.The experimental and testing analysis results show that the proposed method has perfect English semantic translation selection for cross?language information retrieval,its data recall performance is better,and its precision ratio is higher.
作者
方茜
FANG Qian(Chengdu University of Technology,Chengdu 610059,China)
出处
《现代电子技术》
北大核心
2017年第12期39-42,共4页
Modern Electronics Technique
基金
四川省教育厅人文社会科学项目(13SB0090)
关键词
跨语言信息检索
语义翻译
语义选取
语义映射
cross language information retrieval
semantic translation
semantic selection
semantic mapping
