基于查询相关性分析的检索结果聚类算法

发布时间：2018-06-12 02:02

本文选题：检索结果聚类 + 簇描述短语　；参考：《小型微型计算机系统》2011年10期

【摘要】：随着互联网的普及和网页数量的飞速增长,搜索引擎已经成为从网上获取信息的首选工具.然而,目前主流的搜索引擎在响应用户提交的检索请求时,往往以较长的一维列表形式分页展示结果,为了找到自己所需要的信息,用户必须对该结果列表进行耐心的浏览.为了进一步提高用户获取信息的效率和质量,减轻用户的劳动强度,研究者提出了对检索结果进行再挖掘、再组织的问题,聚类就是其中的研究热点之一.本文在分析现有检索结果聚类算法存在的问题的基础上,提出了基于查询相关性分析的标签驱动聚类算法,该算法通过分析短语与查询项的关联程度,提取作为候选簇标签的短语,然后根据这些标签确定网页摘要隶属的候选簇,最后基于对候选簇和标签的评价进行簇筛选和归并,得到聚类结果及每个簇的标签.在相同环境下进行的对比实验表明,所提出的算法优于相关工作,而且需要更少的信息资源支持.
[Abstract]:With the popularity of the Internet and the rapid growth of the number of web pages, search engines have become the first choice to obtain information from the Internet. However, when the mainstream search engines respond to the requests submitted by users, they often display the results in the form of a long one-dimensional list. In order to find the information they need, the users must browse the result list patiently. In order to further improve the efficiency and quality of users' access to information and reduce the labor intensity of users, researchers put forward the problem of remining and reorganizing retrieval results, among which clustering is one of the research hotspots. On the basis of analyzing the problems existing in the existing retrieval result clustering algorithms, this paper proposes a tag-driven clustering algorithm based on query correlation analysis, which analyzes the correlation between phrases and query items. The phrases used as candidate cluster tags are extracted, and then the candidate clusters are determined according to these tags. Finally, the clustering results and the labels of each cluster are obtained by clustering and merging based on the evaluation of candidate clusters and tags. The comparison experiments in the same environment show that the proposed algorithm is superior to the related work and needs less information resources.
【作者单位】：四川大学计算机学院;
【分类号】：G353.1

【参考文献】