文档介绍：<< 信息技术
一种基于智能过滤的 Web 个性化推荐模型
苏玉召
1,2
赵
妍
3
许德山
1,2
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
〔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。
〔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〕G350
TP311
A Model of the Intelligent Filtering for Web Personalization mendation
Su Yuzhao1,2
Zhao Yan3
Xu Deshan1,2
1
2
3
National Science Library,Chinese Academy of Sciences,Beijing 100190
Graduate University of Chinese Academy of Sciences,Beijing 100049
Zhengzhou Institute of Aeronautical Industry Management,Zhengzhou,450015
〔Abstract〕mend system is perform to mend personalized content for user according to the user model and is the key tech-
nology of personalization application,which includes three patterns: rules-based filter system,content-based filter system,collabora-
tion-based filter system and hybrid filter system. The former three mendation 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 ings. It’s excellent feature is that it can both offline learning model an