一种隐私保护机制的POI推荐模型分析  

Analysis of a POI Recommendation Model with a Privacy Protection Mechanism

作　　者：邓玉洁 DENG Yujie(Department of Investigation,Shaanxi Police College,Xi’an 710021,China)

机构地区：[1]陕西警察学院,侦察系,陕西西安710021

出　　处：《微型电脑应用》2024年第12期312-316,共5页Microcomputer Applications

摘　　要：为了保护用户隐私数据和提升位置网络中推荐服务的水平,提出一种基于隐私保护的兴趣点(POI)推荐模型。在原始数据中随机加入隐私保护机制,从而防止攻击者使用推荐结果攻击用户。每个用户的数据通过隐私保护机制处理及遵守相关法律约束后上传到远程服务器,并通过推荐模型分析数据信息。仿真实验表明,当扩展半径为2时,所提隐私保护模型准确率为0.762,较2种比较模型分别提升了0.026和0.009。该模型对位置网络中用户推荐时数据隐私提供了一定借鉴作用。In order to protect users’privacy data and improve the recommendation service level in location network,a point of interests(POI)recommendation model based on privacy protection is proposed.A privacy protection mechanism is randomly added to the original data to prevent attackers from attacking users with recommendation results.The data of each user are processed through the privacy protection mechanism and abided by relevant legal constraints,uploaded to the remote server.The data information is analyzed through the recommendation model.Simulation results show that when the extended radius is 2,the accuracy of the proposed privacy protection model is 0.762,which is 0.026 and 0.009 higher than two comparative models,respectively.It shows that the model provides a reference for data privacy when users recommend in location network.

关 键 词：位置网络 隐私保护 法律约束 社区检测 本地差分隐私 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]