| 摘要: |
| 本文提出了一种面向屈折语的上下文相关自然语言词法分析方法，它依据单词的外部形态和其语法、语义特性及上下文成分的特征，建立了形态、语法和语义之间的联系，使得词法分析的结果准确可靠，并且可以利用语言屈折变化的一致性关系，排除不合理的分析结果，减轻语法分析的负担. |
| 关键词: 机器翻译 词法分析 自然语言处理 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| AN INFLECTED LANGUAGE ORIENTED CONTEXT-SENSITIVE MORPHOLOGICAL ANALYZER |
|
Ke Lun,Huang Heyan,Chen Zhaoxiong
|
| Abstract: |
| This paper presents an inflected language oriented context-sensitive morphological analysis approach. In this approach, not only the appearances of words but also their syntactic and semantic attributes are distinguished. Further more, the context in which a word appears is also considered. Thus a more reliable result will be attained. To relax the burden of syntactical analysis, inflection relations between connected words are used to remove unreasonable results. |
| Key words: Machine translation morphological analysis natural language processing. |
