首页 | 本学科首页   官方微博 | 高级检索  
     

逐步求精法获取上下文无关文法
引用本文:张瑞岭,董韫美.逐步求精法获取上下文无关文法[J].计算机研究与发展,2000,37(1):30-36.
作者姓名:张瑞岭  董韫美
作者单位:中国科学院软件研究所计算机科学开放研究实验室,北京,100080
基金项目:本课题得到国家自然科学基金(项目编号69673018)资助.
摘    要:文法推断研究如何从语言的有限实例,通过归纳推断获取语言的文法定义。文中提出一个基于逐步求精的上下文无关文法推断方法,以尝试将文法推断用于替代或帮助传统手工的文法构造工作。文中的推断方法以Angluinh的交互式学习模型为框架,以逐步求精和复用为主要策略,具有增量式获取结构自然的文法的特点。

关 键 词:归纳学习  上下文无关文法  逐步求精  形式语言

A STEPWISE REFINING METHOD FOR INFERRING CONTEXT-FREE GRAMMARS
ZHANG Rui-Ling,DONG Yun-Mei.A STEPWISE REFINING METHOD FOR INFERRING CONTEXT-FREE GRAMMARS[J].Journal of Computer Research and Development,2000,37(1):30-36.
Authors:ZHANG Rui-Ling  DONG Yun-Mei
Abstract:Grammatical inference(GI) is a problem of learning languages from examples where the target domain is a formal language and the representation class is a family of grammars. A constructive method for inferring context\|free grammars from positive examples is presented in this paper. This work is actuated by the intention to apply GI to replace or supplement the conventional manual grammar construction. The inferring method has the following features: (1) it is interactive; (2) it is incremental; (3) it employs such strategies as stepwise refining and reuse to enhance its learning capability. and (4) the inferred grammars have natural structures.
Keywords:inductive learning  grammatical inference  context\|free grammars  reuse  stepwise refining
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号