一种基于智能过滤的Web个性化推荐模型  

A Model of the Intelligent Filtering for Web Personalization Recommendation

作　　者：苏玉召[1,2] 赵妍[3] 许德山[1,2] 

机构地区：[1]中国科学院文献情报中心,北京100190 [2]中国科学院研究生院,北京100049 [3]郑州航空工业管理学院,郑州450015

出　　处：《图书情报工作》2011年第13期112-115,共4页Library and Information Service

摘　　要：Web个性化研究的关键技术是推荐系统,其作用是根据用户模型推荐个性化内容,当前推荐技术的研究主要包括四种模式:基于规则过滤、基于内容过滤、基于协作过滤和混合过滤模式。前三种工作模式采用的是传统技术和方法,根据当前推荐系统研究的重点和热点,提出一种Web个性化应用的智能过滤推荐模式。智能过滤推荐模式组合采用以上三种工作模式的优点、避免前三种单一模式的缺点。该方法的突出特点是根据离线学习模型提取的用户偏好特征,实现在线智能推荐。Recommend system is perform to recommend personalized content for user according to the user model and is the key technology of personalization application, which includes three patterns： rules-based filter system, content-based filter system, collaboration-based filter system and hybrid filter system. The former three recommendation systems use. General techniques and methods. Now the researchers are focusing on the hybrid methods. The hybrid filtering system gathered the advantages of the former three and avoids their shortcomings. It＇s excellent feature is that it can both offline learning model and online intelligent recommendation.

关 键 词：Web个性化推荐系统 WEB挖掘 基于规则过滤 基于内容过滤 协作过滤 

分 类 号：G202[文化科学—传播学]