基于主题树的个性化元搜索引擎  

Personalized meta search engine based on subject tree

作　　者：张宗仁[1] 杨天奇[1] 

机构地区：[1]暨南大学计算机科学系,广东广州510632

出　　处：《计算机工程与设计》2011年第1期149-152,288,共5页Computer Engineering and Design

基　　金：广东省科技计划基金项目(2007B031300001)

摘　　要：针对元搜索引擎在个性化处理上的不足,提出了主题树的方法,把用户的查询映射到主题树上并返回相关查询,比较准确地表达了不同用户的检索需求。根据用户的浏览行为提出了准确度偏差,用户满意度的评价公式,以用户为中心反映排序是否符合用户的需要。结果排序时不仅考虑了页面的标题、摘要和用户查询的相关性,也考虑了各成员搜索引擎的原来的排序信息以及关键词出现的个数和位置。实验结果表明,该方法有效地提高了元搜索引擎的结果排序能力和个性化处理能力,结果的显示顺序基本符合用户的需要。According to the shortcoming of the meta search engine handing personalization,a subject tree method is proposed.The method maps the user＇s queries to subject tree and return the relevance queries,which can meet more accurately the needs of different users.According to the user＇s browsing behavior,accuracy of deviation and customer satisfaction evaluation of the formula is put forward and reflect the sort of user-centric meets the needs of users.Sort results considers not only the relevance between page＇s title, page＇s abstract and user＇s queries,but also various members search engine＇s original ranking information,as well as the number and location of the Key words appear.The experiment shows that effectively improved the results of meta-search engine capabilities and personalized sequencing capacity,the order accords with user＇s needs basically.

关 键 词：元搜索引擎 主题树 个性化 信息检索 相关度 结果合成 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]