基于用户兴趣的个性化信息检索方法研究  

Personalized Information Retrieval Approach Based on User Interest

作　　者：张一洲[1] 

机构地区：[1]中共淮安市委党校,江苏淮安223003

出　　处：《现代情报》2015年第6期25-28,42,共5页Journal of Modern Information

基　　金：江苏省高校社会科学基金项目(项目编号:No.2012SJD870001);淮安市科技支撑(工业)项目(项目编号:No.HAG2012055)

摘　　要：在信息检索中对基于用户兴趣的检索结果进行重排得到广泛关注。为了构建用户兴趣的知识库,本文对用户的登录细节和点击数据进行综合分析,提出了定制用户访问信息的方法,同时采用开放式目录项目Dmoz自动进行用户兴趣主题映射,对搜索结果进行个性化分类,并根据用户兴趣对检索结果重排,比正常的搜索引擎更容易找到相关的信息。联机实验结果表明,本文提出的方法可有效地提高用户检索精度。Re - ranking of the retrieval results based on the user＇s interests has received wide attention in information re- trieval. In order to build the knowledge base about user＇s interests, the propesed approach to access information was taken into consideration login details and click- through data. This paper automatically mapped Dmoz Open Directory Project topics to users＇ interests, categorized and personalized retrieval results according to user interests and re - ranking of the results was done based on user interests, This made it easy to find relevant document faster than normal search engines. Online experimental results showed that the proposed approach could be effectively used for improving the precision of user retrieval.

关 键 词：信息检索 用户兴趣 主题映射 个性化分类 

分 类 号：G252.7[文化科学—图书馆学]