一种自动获取用户兴趣的个性化搜索算法  

A Personalized Search Algorithm for Automatic Learning of User’s Preference

作　　者：陶剑文[1] 

机构地区：[1]浙江工商职业技术学院计算机应用研究所,浙江宁波315012

出　　处：《情报理论与实践》2007年第4期562-565,共4页Information Studies:Theory & Application

基　　金：浙江省教育厅科研项目资助(20040120);浙江省教育厅青年教师科研基金资助。

摘　　要：解决用户的模糊查询问题一直以来是信息检索领域研究的热点。为了解决不同用户间的查询差异,一种称为个性化搜索的技术得以提出,其通过获取用户的喜好来识别查询意图,但研究发现很少有用户愿意直接或间接提供个人信息。本文提出一种基于用户点击历史信息自动获取用户兴趣进而对搜索结果进行个性化呈现的Web搜索系统架构。基于主题相关PageRank技术,设计了用户兴趣学习算法和个性化搜索页面排序算法。实验表明该算法能有效学习用户的兴趣信息,提高了个性化Web搜索质量。To solve user＇ s ambiguous query problem has long been a hotspot for study in information retrieval field. In order to settle the query difference among different users, a personalized search technology is put forward, which identifies the intention of a query by learning user＇ s preference. However, studies show that few users are willing to provide personal information directly or indirectly. This paper introduces a Web search system model, which learns user＇ s preference automatically based on his click history and provides personalized search results. User＇s preference learning algorithm and personalized page ranking algorithm based on subject-related PageRank technology are introduced. The experiment shows that these algorithms can effectively learn user＇ s preference and improve the quality of personalized Web search.

关 键 词：信息检索 模糊检索 用户兴趣 算法 

分 类 号：G354[文化科学—情报学]