首页 | 本学科首页   官方微博 | 高级检索  
     

基于自然语言理解的受限领域自动应答系统
引用本文:余正涛,樊孝忠,康海燕.基于自然语言理解的受限领域自动应答系统[J].计算机工程,2004,30(18):35-37.
作者姓名:余正涛  樊孝忠  康海燕
作者单位:1. 北京理工大学计算机科学与工程系,北京,100081;昆明理工大学信息工程与自动化学院,昆明,650051
2. 北京理工大学计算机科学与工程系,北京,100081
摘    要:提出了一种在受限领域内进行自动应答的系统总体结构,通过构建领域词汇库、领域本体库、领域真实问句语料库及领域知识文本库等领域资源库,并借助于这些领域资源来理解问句和知识文本,在一定程度上降低了理解的难度。文章以银行业务咨询服务为例,分析了系统结构及实现过程,对问句分析、答案析取、问句语料库构建、文本知识标注等功能模块的实现进行了详细介绍。

关 键 词:受限领域自动应答系统  领域资源库  问句语料库  分词  标注
文章编号:1000-3428(2004)18-0035-03

Restricted Domain Question Answer System Based on Natural Language Understanding
YU Zhengtao,FAN Xiaozhong,KANG Haiyan.Restricted Domain Question Answer System Based on Natural Language Understanding[J].Computer Engineering,2004,30(18):35-37.
Authors:YU Zhengtao    FAN Xiaozhong  KANG Haiyan
Affiliation:YU Zhengtao1,2,FAN Xiaozhong1,KANG Haiyan1
Abstract:This paper proposes an architecture of question answer in restricted domain. It composes domain databases including domain glossary database, domain ontology database, domain question sentence corpus,domain knowledge text database. To understanding the question and knowledge relies on these databases, which reduces the difficulty of understanding in some degree. In this paper, bank counseling is as example. It analyzes the structure and realization of bank counseling, and introduces question analyses, answer extraction, question sentence corpus construction,knowledge text tagging in details.
Keywords:Restricted domain question answer system (QAS)  Domain resource database  Question sentence Corpus  Split  Tagging
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