摘要
针对传统个性化推荐系统存在的隐私容易泄露的缺点,提出一个基于代理的智能推荐系统,在向用户提供准确方便的内容推荐服务的同时保护用户隐私。在该系统中,所有用户私有信息的操作都在客户端执行,使用户隐私得到完善的保护。以嵌于RSS阅读器中的个性化广告系统为例,表明该方法能准确地推荐用户感兴趣的内容并且保护用户隐私。
In order to solve the problem of privacy leak in traditional personalized recommendation systems, this paper presents an Agent-based personalized recommendation system. It offers good-quality recommendation service and protects the user's privacy. As all DUP(Dynamic User Profile)-related operations are working at the client side, the privacy of user will be fully protected. Experiment on an RSS advertising application shows that this system can provide more precise interested messages to users and protect the users' privacy.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第8期283-284,F0003,共3页
Computer Engineering
关键词
个性化推荐系统
隐私保护
数据挖掘
personalized recommendation system
privacy protection
data mining