论文部分内容阅读
【目的】探讨开发汉语可计算语法的理论基础和实践经验。【方法】以HPSG理论为句法框架,以MRS为语义描写手段,以“汉语普通话在线语法”(简称“汉构”)的开发过程为例,重点研究通过构建词库和规则层级,对汉语特殊结构进行计算实现。【结果】“汉构”的开发证明,HPSG非常适合作为汉语可计算语法开发的理论框架。【局限】“汉构”仍在继续开发中,其覆盖率暂未接受大规模自然语料的检测。【结论】“汉构”可作为沟通形式语言学和计算语言学的桥梁和开发大规模资源型语法的基础。
【Objective】 To explore the theoretical basis and practical experience of developing Chinese computable grammar. 【Method】 Taking the HPSG theory as the syntactic framework and MRS as the semantic description method, taking the development process of “Chinese Mandarin Online Grammar ” (abbreviation “Han Cao ”) as an example, this paper focuses on the construction of thesaurus and rules hierarchy , Calculated on the Chinese special structure. 【Result】 The development of “Han Architecture” proves that HPSG is very suitable as a theoretical framework for the development of Chinese computable grammar. [Limitations] “Han structure ” is still under development, its coverage has yet to accept the large-scale detection of natural corpus. 【Conclusion】 “Han Cao ” can be used as a bridge between formal linguistics and computational linguistics and as a basis for developing large-scale resource-based grammars.