摘要
分析大陆和台湾的在农业方面的语言差异、简繁体的编码问题和两岸文字的语义差别,在此基础上提出两岸文字转化的规则。研究中文分词、多目标字词消歧和人工辅助转换等主要转换技术,重点研究语境模糊匹配消歧,即设立语境库并通过语境模糊匹配算法,来减少单字转换时出现的歧义现象。采用双数组Trie树的索引结构和逆向最大匹配的分词算法、KMP模式匹配算法等,提出面向两岸农产品物流信息的汉字转换系统设计方案,并在基于物联网的海峡两岸农产品物流商务平台中予以实现,取得较好效果。
This paper presented some rules for cross-strait text translation based on the language difference, simplified and traditional character encoding difference and semantic difference in agriculture. Some main text conversion techniques, such as Chinese word segmentation, multi-target word disambiguation, and artificial auxiliary conversion, were studied. The focused research was contextual fuzzy matching disambiguation, namely setting up context library and reducing the ambiguity phenomenon through the contextual fuzzy matching algorithm when a separate word was converted. The Double-Array Trie structure, reverse maximum matching segmentation algorithm, and KMP pattern matching algorithm, were applied to design the Chinese character conversion system for cross-strait agricultural products logistics information. The proposed scheme has been implemented and tested in the business platform of cross-strait agricultural products logistics based on IOT. The results showed good performance of the designed translation engine in practice.
出处
《农业网络信息》
2014年第7期5-9,64,共6页
Agriculture Network Information
基金
国家科技支撑计划(编号:2012BAH21F04)
关键词
海峡物流
信息翻译
汉字转换
农产品
中文分词
cross-strait logistics
information translation
Chinese character conversion
agricultural product
Chinese segmentation
