中文文法规则的化简及其语义计算效果对比

摘要

为在自然语言处理中进行更快的语义计算，分析自然语言中最普适语法规则的基本特征，在此基础上，将语义功能相似的文法符号合并，构造一个更为简洁的文法规则，用来描述自然语言；依据自然语言语义计算模型，通过语义计算，进行模型求解，求出最符合语义逻辑的语法分析方案。通过实验比较文法简化前后的计算复杂度和正确率，得到模型求解“时间复杂度”和“正确率”之间的关系。%To make faster semantic computation in natural language processing,many grammar symbols which were similar in semantic function were merged to one grammar symbol according to the basic characteristics of the natural language,a more con-cise grammar rules was constructed to descript the Chinese natural language,and the best syntax analysis method in semantic logical was achieved and the semantic model of the natural language was solved by semantic computing.Finally through experi-ments,the computational complexity and correct rates of the new grammar rules were compared with the old grammar rules,and the relations between the computational complexity and the correct rates in the process of solving the semantic model were searched out.