面向隐私保护推荐系统的安全两方协议研究  

ON SECURE TWO-PARTY PROTOCOL FOR PRIVACY PROTECTION RECOMMENDATION SYSTEM

作　　者：刘洪伟[1] 梁飞[1] 朱慧[1] 

机构地区：[1]广东工业大学管理学院,广东广州510520

出　　处：《计算机应用与软件》2014年第8期17-19,共3页Computer Applications and Software

基　　金：国家自然科学基金项目(70971027)

摘　　要：在网络购物如火如荼的今天,为了更具竞争力,两家彼此竞争的公司也可能在某些领域进行合作。比如,两家电子商务网站为了给用户提供更好的推荐服务,传统的做法是双方共享对方的评分数据库,但是这种做法显然没能保护双方数据的隐私性。在UICF推荐算法的基础上设计一个安全两方协议,有效地解决了隐私保护协同过滤推荐中的两方协同计算问题。In today the online shopping is getting in full swing,two competing companies may also cooperate in certain areas in order to be more competitive. For example,in order to provide users with better recommendation,the traditional approach is that two e-commerce sites share the ratings database each other,but this is clear not able to protect data＇s privacy of both. We design a secure two-party protocol on the basis of UICF recommendation algorithm,which effectively solves the collaborative two-party computation problem in privacy protection collaborative filtration and recommendation.

关 键 词：推荐算法 协同过滤 隐私保护 安全两方 同态加密 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]