语料库与知识获取模型  

CORPUS&KNOWLEDGE ACQUISITION MODEL

作　　者：张敏[1] 罗振声[1] 

机构地区：[1]清华大学

出　　处：《中文信息学报》1994年第1期15-24,共10页Journal of Chinese Information Processing

摘　　要：在计算语言学中，知识作为句法分析的数据支持起着重要的作用，如何利用计算机从现实世界中获取知识一直是人工智能领域探讨的重要课题，本文描述了如何从语料库这一客观真实数据资料获取有用知识，用以帮助汉语句子中依存关系的分析，在依存语法作为语言模型的基础上，探讨了向上依存关系的二元同现矩阵，以及知识库的组成和结构，在获取词跟词之间的依存关系的同时，还获取了词跟类、类跟类之间的依存关系。Abstract Knowledge which supports syntatic analysis is overiding important in Computational Linguistics. How to make full use of computer to acquire knowledge form the objectiveworld is one of the most significant parts in the field of Artificial Intellingence. This paper discusses how to acquire knowledge from corpus and apply this knowledge to Chinese dependency parsing. Taking Dependency Grammar as language processing model, this thesis approaches the binary synchronous matrix of upward dependency as well as the component and the structure of knowledge base. Word to word dependency, word to class dependencyand class to class dependency are all obtained.

关 键 词：语料库 知识获取模型 人工智能 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]