面向主题的概念检索研究  

Research on Conceptual Information Retrieval Faced to Subject

作　　者：李小兵[1] 杨一平[2] 刘曙光[1] 郭域峰[1] 

机构地区：[1]中国科学院研究生院,北京100039 [2]中国科学院自动化研究所综合信息系统研究中心,北京100080

出　　处：《计算机工程与应用》2005年第34期168-171,共4页Computer Engineering and Applications

摘　　要：该文提出了一种基于概念网络和主题概念树的面向主题的文本检索算法。依托概念网络建立主题概念树,利用主题概念树对用户的查询请求进行语义扩展,实现同义和语义蕴涵检索。关联度的计算模型考虑了词与词之间,句与句之间的语义激励。通过关联度在主题概念树上的传播模型,实现复合概念关联度的计算。检索结果按关联度大小降序排列。基于主题概念树的概念检索导航为用户检索提供了便利。This paper proposes a text retrieval algorithm based on conceptual network and subject concept tree.The subject concept tree is built on the conceptual network,and it is used to make semantic expansion to the user＇s information requirement,synonymy and semantic implication information retrieval are achieved through this expansion.The computation of association degree considers the semantic activation between the words and between the sentences. Association degree of compound concept can be received through the association degree＇s spread model of subject concept tree.The retrieval resuhs are presented in the association degree＇s descending order, Conceptual information retrieval navigation based on the subject concept tree provides convenience to the user.

关 键 词：概念检索 关联度 概念网络 主题概念树 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]