基于HITS算法的查询结果多样化方法  

The search result diversification approach based on the HITS algorithm

作　　者：陈飞[1] 张敏[1] 刘奕群[1] 马少平[1] 

机构地区：[1]智能技术与系统国家重点实验室清华大学信息科学与技术国家实验室清华大学计算机系,北京100084

出　　处：《山东大学学报（理学版）》2011年第5期44-48,共5页Journal of Shandong University(Natural Science)

基　　金：国家自然科学基金资助项目(60736044,60903107);高等学校博士学科点专项科研基金资助项目(20090002120005)

摘　　要：现有的查询结果多样化研究很难准确得到用户多样性需求并提供与用户查询各个方面需求相关的文档。针对这个问题,本文基于HITS算法的网页间链接分析特性,根据网页链接图直接计算查询结果列表中的文档可能满足用户多样性需求的程度,并将其应用到结果列表的重排序中以实现搜索结果多样性。在TREC大规模数据集合上的实验结果表明了该方法的有效性。To avoid the problem that users′ diversity needs cannot be precisely obtained or documents provided cannot concern all aspects of the needs in a specific query,a new method was proposed based on the link-parsing feature of the HITS algorithm,in where the possibility was directly calculated according to the diversity of documents in the search result list for a query,and then the result list was re-ranked based on this value.Experimental results on the TREC′s large-scale data collections verified that this method was effective.

关 键 词：多样性 HITS PAGERANK 权威性 中心性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]