一种基于主题树模型的Web主题检索算法  

A Web Topic-specific Information Search Algorithm Based on Topic-specific Tree Model

作　　者：杨创新[1] 彭宏[1] 杨沛[1] 

机构地区：[1]广东商学院网络中心,广州510640

出　　处：《计算机工程与应用》2004年第16期184-187,226,共5页Computer Engineering and Applications

摘　　要：面向主题的Web检索是当前的一个研究热点。而常规的关键词检索方式的主要缺点是不能准确地表达用户的查询意图。为此,文章在概念语义空间的基础上构建了一个新的主题树模型,并提出了一种基于主题树模型的主题扩展算法,可应用于主题检索;同时,提出了主题偏移度和切题度等概念,用于动态跟踪用户的检索过程以进行查询反馈。实验表明,主题树模型及相关的检索算法,能对主题检索进行有效地扩展,减少用户检索过程中产生的主题漂移,提高检索的速度和查准率。The topic-specific Web information search is one of hotspot being studied at present.By the ordinary key-word information search method,user's can't express their intention accurately.So we build a new model of topic-specific tree,and give the algorithm for topic-specific expansion,which can be applied to topic -specific search.At the same time ,there are two conceptions,″Topic offset″and″Approach to Topic″,would give users feedback for tracing their search process.The experiment shows ,the topic-specific tree and it's correlative algorithm,would expand the search words effi-ciency,reduce the topic-specific excursion accelerate process and improve precison in the information search process.

关 键 词：主题树 概念语义空间 信息检索 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]