论文部分内容阅读
面对基于双语词典的跨语言检索查询翻译方法中固有的一对多等翻译模糊问题,已有研究成果存在对于非组合型复合词无法进行准确翻译、双语词典和其他翻译资源联合使用引入较大计算开销等弊端。为建立英汉双向跨语言检索实用性系统,在现有的一部包含若干科技词汇和短语的双语科技词典的基础上,着重研究如何引入平行语料来改进已有的双语词典问题。目标是生成一部基于句对齐平行语料的科技类双语概率词典,为跨语言检索查询翻译消歧提供实时性支持。
In the face of the inherent one-to-many translation ambiguity problems in cross-lingual search query translation methods based on bilingual lexicons, the existing research results can not accurately translate non-combined compound words. The combination of bilingual lexicons and other translation resources introduces a large calculation Expenses and other drawbacks. In order to establish a practical system of bi-directional cross-linguistic search in English and Chinese, this paper focuses on how to introduce parallel corpus to improve the existing bilingual dictionary based on the existing bilingual scientific dictionary containing several scientific and technical terms and phrases. The goal is to generate a bilingual scientific bilingual dictionary based on sentence aligned parallel corpus to provide real-time support for cross-language retrieval query disambiguation.