网络环境中基于语义聚类的多关键词查询机制  

A Multi-keyword Query Mechanism Based on Semantic Cluster in Network Environments

作　　者：海沫[1] 郭树行[1] 

机构地区：[1]中央财经大学信息学院,北京100081

出　　处：《图书情报工作》2012年第20期37-41,共5页Library and Information Service

基　　金：<图书情报工作>杂志社2011年度出版基金项目"面向网络科技文献资源的知识抽取及其语义关联的可计算分析"(项目编号:2011CB002)研究成果之一

摘　　要：为提高多关键词查询的效率并减少多关键词查询的开销,提出一种基于语义聚类的多关键词查询算法——MKQBSC。该算法使得语义相似的节点聚为一类,节点加入、退出或节点的语义改变时,聚类将相应改变。查询请求在相邻的语义聚类之间转发,直至到达语义相似的聚类。仿真实验结果表明:与传统的基于对倒排表求交集的多关键词查询算法相比,MKQBSC算法所需的路由跳数和所产生的消息数更少。In order to increase the efficiency as well as decrease the cost of muhi-keyword query, a multi-keyword query algorithm based on semantic duster-MKQBSC is proposed. This algorithm clusters peers with similar semantics. When any peer joins, leaves or the semantics of any peer changes, peer cluster will change. A query request keeps being forwarded among neighbor semantic clusters until it arrives at a cluster with similar semantics. Simulation results show that compared with traditional multi-keyword query algorithm based on the intersection of inverted lists, both the routing hops and number of messages of our MKQBSC algorithm are less.

关 键 词：多关键词查询 语义聚类 路由跳数 消息数 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]