首页> 外国专利> Method and system for privacy protection recommended against contributes scored users based on matrix factorization

Method and system for privacy protection recommended against contributes scored users based on matrix factorization

机译:基于矩阵分解的针对贡献评分用户推荐的隐私保护方法和系统

摘要

A method and system for securely generating a recommendation by matrix factorization, without having to learn the contents of any individual record, receive a set of records including a set of pairs and entry token as input a step, to design and evaluate Gaburu circuit based on matrix factorization for a set of records, any person records, or includes a recommendation, without even learning content of any information extracted from the record, at least one One of about a way of privacy protection item, start by generating a recommendation. The system includes: a plurality of users or database represents the record source of; includes three of the party consisting of a recommender system to evaluate the cryptographic service provider and the circuit to design a Gaburu circuit, as a result, have been extracted from the record and the record any of the information is kept secret to the party other than these sources, the recommendation is only known by the requesting party.
机译:一种用于通过矩阵分解安全地生成推荐的方法和系统,而无需学习任何单个记录的内容,而是接收包括一组对和输入令牌的一组记录作为输入步骤,从而基于以下步骤设计和评估Gaburu电路对于一组记录,任何人记录或包括推荐的矩阵分解,甚至不学习从记录中提取的任何信息的内容,关于隐私保护项目的一种方式中的至少一种,是从生成推荐开始。该系统包括:代表多个用户或数据库的记录源;包括三个参与方,其中包括一个用于评估密码服务提供商的推荐系统和一个用于设计Gaburu电路的电路,因此,已从记录中提取了该记录,并且该记录中的任何信息均已被该方保密,而不是这些来源中,推荐仅由请求方知道。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号