摘要
提出汉语语义分析的方法,具体策略是借用依存语法作为表示框架,通过在基本内部—外部算法中结合知识识别句子中的依存关系。实验结果说明,利用借助知识的内部—外部算法训练无指导模型是可行的汉语语义分析方法。
This paper presented an approach of semantic analysis on Chinese, which borrowing the dependency grammar as framework of representation. Through combining knowledge with original inside-outside algorithm, it identified the dependency relationships of sentences. The experiments show it is applicable to train the unsupervised model of Chinese semantic analysis with the inside-outside algorithm.
出处
《计算机应用研究》
CSCD
北大核心
2010年第2期529-531,共3页
Application Research of Computers
关键词
依存分析
内-外算法
无指导学习
semantic analysis
inside-outside algorithm
unsupervised learning