英汉机器翻译中的歧义和消歧研究

发布时间：2018-06-12 16:38

本文选题：英汉机器翻译 + 英文歧义　；参考：《河南工业大学》2013年硕士论文

【摘要】：机器翻译在经历了六十多年的发展后，其翻译质量依然没有取得突破性的进展。其中，“歧义”依然是一个大问题。鉴于歧义的不同划分标准，以及作者分析讨论的方便，本文中所讨论的歧义主要着眼于词汇歧义、句法歧义、语义歧义和语用歧义。因此，本文的一个目的就是从纯语言学的角度来探寻英汉文本中的歧义现象，以期对英汉机器翻译中所涉英、汉两语言领域中歧义的不同类型提供一个概括性的汇总。自然语言中歧义现象的存在有着十分复杂的原因，任何人企图用一个简单的公式或原理来完全消除歧义都是行不通的。考虑到机器翻译的一个重要目的就是为使用者提供信息，因此，我们将采用目的论（Skopostheorie，即翻译过程中的最主要因素是整体翻译行为的目的）作为消歧的理论基础；译入语和原语的“适当性或恰当性而不是对等性或精确性”作为消歧的原则。基于此构想，本文主要着眼在英汉文本语言中歧义的相似性，综合并重点讨论了词汇消歧和句法消歧。对于词汇消歧，我们试探性地提出了语料库策略，知识策略以及统计学策略；而对于句法歧义，我们则探讨了规则策略和混合策略。围绕着对英汉机器翻译中歧义和消歧的研究主题，我们采取了一种综合的方法，即定性研究和抽样举例，来收集数据。定性研究致力于回答英汉机器翻译中存在着什么类型的歧义以及如何消歧的问题；而抽样举例则是为了给歧义类型和消歧策略的分析提供代表性的例子。通过该研究，我们发现英汉歧义都是普遍存在且很复杂的一种现象，而英汉机器翻译系统在处理该问题时依然还有很多局限性以及诸多不如意之处。但鉴于歧义的复杂性，消歧之路注定了是曲折和漫长的。总之，，在英汉机器翻译中，无论是英语歧义和消歧，还是汉语歧义和消歧，都还需要做更多的研究、投入更多的时间和精力。
[Abstract]:After more than 60 years of development, the quality of machine translation has not made a breakthrough. Among them, ambiguity is still a big problem. In view of the different criteria of ambiguity and the convenience of the author's analysis and discussion, the ambiguity discussed in this paper is mainly focused on lexical ambiguity, syntactic ambiguity, semantic ambiguity and pragmatic ambiguity. Therefore, one of the purposes of this thesis is to explore the ambiguity in English and Chinese texts from the perspective of pure linguistics, in order to provide a general summary of the different types of ambiguity in the field of English and Chinese machine translation. The existence of ambiguity in natural language has very complicated reasons. It is impossible for anyone to attempt to eliminate ambiguity completely by a simple formula or principle. Considering that one of the important purposes of machine translation is to provide users with information, we will use Skopostheorieas Skopostheorie, which is the most important factor in the process of translation, as the theoretical basis of disambiguation. The principle of disambiguation is "appropriateness or appropriateness rather than reciprocity or accuracy". Based on this idea, this paper focuses on the similarity of ambiguity in English and Chinese texts, synthesizing and focusing on lexical disambiguation and syntactic disambiguation. For lexical disambiguation, we tentatively propose corpus strategy, knowledge strategy and statistical strategy, while for syntactic ambiguity, we discuss rule strategy and mixed strategy. Focusing on the topic of ambiguity and disambiguation in English-Chinese machine translation, we adopt a comprehensive approach, namely qualitative research and sampling examples, to collect data. The qualitative study is devoted to answering the questions of what types of ambiguities and how to disambiguate in English-Chinese machine translation, while sampling examples are intended to provide representative examples for the analysis of ambiguity types and disambiguation strategies. Through this study, we find that ambiguity between English and Chinese is a common and complex phenomenon, and there are still many limitations and many disappointments in the process of dealing with this problem in the English-Chinese machine translation system. But in view of the complexity of ambiguity, the road of disambiguation is doomed to be tortuous and long. In short, in English and Chinese machine translation, both English ambiguity and disambiguation, as well as Chinese ambiguity and disambiguation, need more research and more time and effort.
【学位授予单位】：河南工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：H315.9;H085

【共引文献】