基于动态知识库搜索引擎的技术  

Research on the techniques of search engines based on dynamic knowledge bases

作　　者：王会进[1] 陈超华[1] 李清[1] 

机构地区：[1]暨南大学计算机科学系,广东广州510632

出　　处：《暨南大学学报（自然科学与医学版）》2004年第1期36-40,50,共6页Journal of Jinan University(Natural Science & Medicine Edition)

摘　　要：　"词的不匹配"是全文信息检索中存在的一个基本问题.为解决此问题,已提出过一些查询扩展方法.现提出一种新的基于动态知识库的搜索引擎原型———DKIRS检索系统.它利用用户检索的结果及用户的反馈信息动态地构造知识库,然后基于知识库对初始查询进行扩展,再利用扩展后的查询进行信息检索.A fundamental problem for Information Retrieval (IR) is the 'word mismatch'. Some query expansion approaches have been proposed to solve the problem. A new method based on the dynamic knowledge base is proposed to make the query expansion. A prototype of the Dynamic Knowledge base Information Retrieval System (DKIRS) is implemented and it constructs a knowledge base using the searched results and user's feedback information. Then the initial query is expanded by means of the knowledge base. Finally, the searche information are obtained by the expanded query.

关 键 词：动态知识库 查询扩展 词的不匹配 信息检索 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]