基于Web的中文开放式问题回答系统  

A Web-based Chinese Open-domain Question Answering System

作　　者：林旭东[1] 彭宏[1] 郑启伦[1] 陈绍坚[1] 

机构地区：[1]华南理工大学计算机科学与工程学院,广州510640

出　　处：《计算机科学》2006年第5期211-213,226,共4页Computer Science

基　　金：广东省科技攻关项目(A10202001);广州市科技攻关项目(2004Z2-D0091)

摘　　要：互联网正逐渐成为重要的信息资源,然而大多数搜索引擎不能处理自然语言提出的问题。基于互联网的中文问题回答系统由问题处理、信息检索、答案抽取和答案判断组成,利用命名实体识别、语义依存关系和案例规则模板实现答案抽取。实验表明:命名实体识别、语义依存关系和案例规则模板能有效地实现答案抽取,获得较高正确率。The Internet is increasingly being used as a source of reference information. Most popular search engines, however, are not designed for answering natural language questions. A Web-based Chinese question answering system is made of four parts： question processing, information retrieval, answer extraction and answer justification. It utilizes named entity recognition, semantic dependency relations and case-based rule to realize answer extraction. Experiments show that named entity recognition, semantic dependency relations and case-based rule perform very well in answer extraction.

关 键 词：问题回答 语义依存关系 命名实体识别 信息抽取 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]