本研究基于自建的专门自然语言语料库,甄别出新的量化名词短语的约束条件,并根据计算实现的需要,对Kunoetal.(1999)和杨春雷(2004)提出的约束条件系统进行修改和完善,建立了一个面向量化名词短语语用解释消歧的约束条件系统。研究表明该约束条件系统能区分75%的量化名词短语的指代和非指代语用解释,根据语用解释的细类,点数和的分布显示出较高的集聚性特征。 Based on the self-built specialized natural language corpus, this study identifies the constraints of the new quantified noun phrases and modifies and perfects the system of constraints proposed by Kunoetal. (1999) and Yang Chunlei (2004) , A system of constraint conditions for pragmatic interpretation of disambiguation is established. The research shows that the system of constraints can distinguish 75% of the quantifier noun phrases from the non-idiom explanation, and according to the breakdown of pragmatic explanation, the distribution of the points and the scores shows a high clustering characteristic.