摘要
提出一种方法,用以实现自然语言理解系统的自动词汇扩展,以模拟人的词汇学习能力.该方法的基本要点是:(1)用语法制导切词实现生词的自动辨别;(2)用反复试探句法分析来确定生词的句法功能;(3)简化词典的词义表示形式,使词义隐含在知识库系统中.
This paper presents an approach to the automatic enrichment of lexicon in a natural language understanding system to simulate man's word learning ability. The following is fundamental in this approach: (1) to identify new words by parser-monitored word-by-word segmentation; (2)to sense the syntactic function of new words by trial-and-error parsing; (3) to adopt a simplified word meaning expression in the dictionary entry and have word meaning embodied in the knowledge base system.
出处
《山东大学学报(自然科学版)》
CSCD
1991年第4期409-416,共8页
Journal of Shandong University(Natural Science Edition)
关键词
自然语言
理解
机器词典
自动切词
machine learning
lexicon
automatic word-by-word segmentation
syntactic analysis
knowledge expression