计算机理解汉语需要语法理论支撑  

Natural language understanding needs support from syntactic theory

作　　者：靳光瑾 

机构地区：[1]国家语委

出　　处：《语言文字应用》1999年第2期84-89,共6页Applied Linguistics

摘　　要：自然语言理解的核心问题在于语义理解。首先要求语义解释是形式化的，不仅人能看得懂，而且在机器上是可操作、可计算的。形式化语义研究的首要内容是将自然语言语句、语段、篇章转换成高阶逻辑公式集或者其他等价形式（转换规则与语言有关）。本文研究将汉语语句转换成逻辑公式的过程中需要哪些汉语语法理论支撑。Abstract The key problem in natural language understanding is semantic understanding. In order to make computer understand natural language, the first requirement is that semantic interpretation must be formalized, so that it can not only be understood by human, but also be operated and computed by computer. The most important thing in formalizing semantics is translating sentences, text spans, discourses in natural language into higher order logic formulas (The translation rules are language-dependent). This paper discusses what kinds of support from the Chinese syntactic theory are needed in the translation from Chinese sentences to logic formula.

关 键 词：计算机理解 语法理论 省略句 自然语言理解 计算机科学 现代汉语 生成语法 形式化 “都” 汉语语法 

分 类 号：H085[语言文字—语言学] H085