Translating OOV Phrases Based on Lexical Information and Web Mining

摘要：

This paper presents a novel approach to improvethe OOV(Out-of-vocabulary)phrase translation bycombining lexical information and web mining.Wefirst retrieve the top relevant anchor words with sourceOOV phrase from search engine,and then search thetranslation with expanded query source OOV phrase+ anchor word from mixed-language web pages.Finally,a ME(Maximum Entropy)model is employedto rank returned translation candidates by combininglexical similarity and statistical similarity.Experimentresults show our approach is promising.

作者: GuihuaSun GaopanXu KeZhang

作者单位: Department of Computer,Xiamen University of Technology,Xiamen 361024,China

会议类型: 国际会议

会议名称: 2008 3rd International Conference on Intelligent System and Knowledge Engineering(第三届智能系统与知识工程国际会议)(ISKE 2008)

会议地点: 厦门

会议语种:英文

页码: 791-796

在线出版日期: 2008-11-17（万方平台首次上网日期，不代表论文的发表时间）