一种结合K-Means的层次化的搜索结果聚类方法  

A hierarchical search results clustering method based on K-Means

作　　者：于洪[1] 谌强[1] 

机构地区：[1]重庆邮电大学计算机科学与技术研究所,重庆400065

出　　处：《重庆邮电大学学报（自然科学版）》2010年第3期355-359,共5页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：国家自然科学基金(60773113);重庆市科委项目(CSTC;2009BB2082);重庆市教委项目(KJ080510)~~

摘　　要：针对用户在搜索结果列表中寻找所需信息困难的问题,在分析了Web搜索结果的特点的基础上,提出了一种结合K-Means的层次化方法对搜索结果进行聚类,并通过向用户提供查询结果的类别标签分类显示结果,从而大大提高可浏览性。同时,在该方法的基础上设计并实现了一个搜索结果聚类原型系统,实验结果表明新方法是可行的。Against the difficulties of users in the search results list,a hierarchical search results clustering method based on K-Means was proposed based on the analysis of the features of web search results.The browsability of the search results was greatly improved by providing users with category labels of search results for the classification display.Besides,a prototype clustering search results system was designed and implemented based on this method.The experiment results show that the new method is feasible.

关 键 词：搜索结果聚类 层次化聚类 类标签 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]