Show simple item record

dc.contributor.advisor史晓东
dc.contributor.author刘智文
dc.date.accessioned2016-02-14T08:19:47Z
dc.date.available2016-02-14T08:19:47Z
dc.date.issued2010-12-29 15:44:30.0
dc.identifier.urihttps://dspace.xmu.edu.cn/handle/2288/50866
dc.description.abstract基于统计的机器翻译方法是目前机器翻译领域主流的研究方法。现存的统计机器翻译模型大体可以分为基于词的模型、基于短语的模型和基于句法的模型三大类。其中基于词的翻译模型提出的时间最早,其数学描述比较严密,但是由于在翻译过程中以词作为基本的翻译单位,难以使用语言的上下文信息,故而在词义消歧、语序调整等方面都存在很大的不足。基于短语的翻译模型以短语为基本翻译单位,可以有效利用语言的上下文信息。基于短语的模型也有缺乏全局信息,远距离调序困难等缺点。基于句法的模型从理论上来讲能够利用语言更深层次的信息,具有更大的潜力,但该类模型目前还没有取得人们预期的突破性进展。短语模型是一种十分健壮的模型,模型本身对训练...
dc.description.abstractStatistical machine translation method is the mainstream research method in the field of machine translation now. Existing statistical machine translation models can be roughly divided into three categories, including the word-based models, the phrases-based models and the syntax-based models. The first proposed model is the word-based translation model, which has strict mathematical description. ...
dc.language.isozh_CN
dc.relation.urihttps://catalog.xmu.edu.cn/opac/openlink.php?strText=26013&doctype=ALL&strSearchType=callno
dc.source.urihttps://etd.xmu.edu.cn/detail.asp?serial=27665
dc.subject对齐模板
dc.subject双语词聚类
dc.subject词聚类替换
dc.subject藏文分词
dc.subject条件随机场
dc.subjectalignment template
dc.subjectbilingual word clustering
dc.subjectword clustering replacement
dc.subjectTibetan word segmentation
dc.subjectconditional random field
dc.title藏汉统计机器翻译研究
dc.title.alternativeTibetan-Chinese Statistical Machine Translation Research
dc.typethesis
dc.date.replied2010-05-29
dc.description.note学位:工学硕士
dc.description.note院系专业:信息科学与技术学院计算机科学系_计算机应用技术
dc.description.note学号:23020071151284


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record