文化大學機構典藏 CCUR:Item 987654321/19674
English  |  正體中文  |  简体中文  |  全文笔数/总笔数 : 47249/51115 (92%)
造访人次 : 14073341      在线人数 : 300
RC Version 6.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
搜寻范围 查询小技巧:
  • 您可在西文检索词汇前后加上"双引号",以获取较精准的检索结果
  • 若欲以作者姓名搜寻,建议至进阶搜寻限定作者字段,可获得较完整数据
  • 进阶搜寻
    主页登入上传说明关于CCUR管理 到手机版


    jsp.display-item.identifier=請使用永久網址來引用或連結此文件: https://irlib.pccu.edu.tw/handle/987654321/19674


    题名: 建構中文詞關聯本體論以改善英譯中選詞之研究
    作者: 黃珮茹
    贡献者: 資訊管理學系
    关键词: 英譯中
    本體論
    詞彙選取
    知識庫
    日期: 2009
    上传时间: 2011-10-04 14:43:33 (UTC+8)
    摘要: 不同語系的語言,在表達同一語意所使用的語法及句型結構常有不同,習慣用語也不盡相符,同時,一字(或詞)多義也造成兩種語言翻譯時選詞的謬誤,諸多因素形成翻譯的障礙,使翻譯成為高門檻的工作。
    機器做語言間的自動翻譯已盛行多年,但一字(或詞)多義、句型及時態和比對等因素,使機器翻譯一直無法令使用者滿意。本研究以英譯中時的中文選詞進行研究,提出建構中文詞關聯的本體論,運用詞關聯可提升英譯中時的選詞正確率。
    詞的意義往往和其前、後詞及在句中所扮演的詞性和地位相關,正確的選詞必須考量詞的關聯性,故在本研究中,以文句中較常出現且相對重要的詞性如名詞、動詞、形容詞及副詞等,建立其間常用的五個法則,並透過這些法則,分析給定的中文文句,建構中文詞關聯本體論,以展現中文詞關聯知識。
    五個法則包括名詞+名詞、形容詞+名詞、副詞+動詞、名詞+動詞、動詞+名詞,遞迴運用此五個法則,可以分析出英文字的適當中文譯義。根據實作驗證,給定建構中文詞關聯本體論的中文文句數愈多,愈可以改善選詞的狀況,提高選詞正確率。
    Translation between languages is a high challenge work due to difficulties occurred during the translation process. Mapping and analysis between sentence structures, proper semantic selection of polysemous words, translation of idioms, and so on are major factors causing difficulties in translation between different language systems.
    Because of the above factors, machine language translators cannot meet users’ satisfaction even thought they have been published for years. The research emphasizes on Chinese word selection in English-to-Chinese translation, especially on polysemous words. The associated relationships among Chinese words are discovered to construct an ontology. The Chinese word-relationship ontology is capable of improve the accuracy in English-to-Chinese translation.
    The exact meaning of a word is determined on its part of speech and its associated pre-words and post-words in the sentence. In this thesis research, five rules which contain important and meaningful part of speech such as verb, noun, adjective, adverb, are established to detect word relationships and word status of translating language in a sentence. They are (R1) noun + noun; (R2) adjective + noun; (R3) adverb + verb; (R4) noun + verb, and (R5) verb + noun. By recursively applying these rules along with the knowledge in the Chinese word-relationship ontology, the Chinese translation of the English word is determined.
    Our experiments showed that use of the ontology and increase the number of rules in the ontology can improve the accuracy of word selection obviously.
    显示于类别:[資訊管理學系暨資訊管理研究所 ] 博碩士論文

    文件中的档案:

    档案 描述 大小格式浏览次数
    index.html0KbHTML170检视/开启


    在CCUR中所有的数据项都受到原著作权保护.


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - 回馈