基于概念聚类的Web数据挖掘搜索引擎的设计与实现  

Search Engine Design and Implementation of Web Data Mining Based on Concept Clustering

作　　者：刘典型[1,2] 刘完芳[1] 钟钢[1] 

机构地区：[1]湖南警察学院信息技术系,湖南长沙410138 [2]湖南大学软件学院,湖南长沙410082

出　　处：《软件工程师》2015年第5期18-20,共3页Software Engineer

基　　金：湖南省科技创新基金项目资助(项目编号:2014FJ3108)

摘　　要：针对Web数据挖掘的搜索过程,其准确度很大程度取决于用户输入的关键词的数量,以及搜索引擎对关键词的语义的解析与用户原意的吻合度,而搜索引擎对关键词的解析,包括基于链接的聚类方法和基于概念的聚类方法。本文克服基于链接的聚类方法的缺陷,采用基于概念聚类的方法,从二分图的概念和存储方法入手,设计和实现了个性化的Web数据挖掘搜索引擎,并验证了其优越性。For the search of the web data mining,its accuracy depends to the numbers of keywords that user has inputted very much,as well as the agreement of user’s intent and the semantic analysis by search engine,the analysis of the search engine including clustering method based on link and based on concept.In this paper,to overcome the defect of the clustering method based on link,through using the method of clustering based on concept,starting from the concept and the storage method of bipartite graph,designed and implemented a personalized search engine of web data mining,its superiority is verified.

关 键 词：二分图 邻接矩阵 聚类 数据挖掘 搜索引擎 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]