本地化差分隐私在数据众包中的应用

Local differential privacy applications in data crowdsourcing

近年来,众包技术受到产、学界广泛关注。采用众包模式,某些机构可以在获得用户同意后,使用用户的行为模式数据、空间位置数据、主动提交数据等进行分析,或者将任务通过众包平台发布,借助大众的合力来完成。现实生活中存在着大量类似的问题,因此众包有着广大的应用前景。然而,尽管众包已为许多行业和产品带来极大效益,但众包过程中涉及的隐私信息泄露问题尚未得到很好的解决。本地化差分隐私保护技术由于其能保证较强的隐私保护能力与较高的数据可用性,成为当前众包隐私保护的热门技术。

In recent years,data crowdsourcing technology has received extensive attention from the industry and academia. With the data crowdsourcing model,some organizations can use the user＇s behavioral pattern data,spatial location data,and active submission data for analysis,or perform tasks by publishing them on a crowd-sourced platform. Data crowdsourcing has a broad application prospects because of the huge amount of similar problems in real life. However,although data crowdsourcing has brought great benefits to many industries and products,the disclosure of privacy information involved in crowdsourcing has not been well resolved. Local differential privacy has become a popular technology for crowdsourcing privacy protection because of its ability to ensure strong privacy protection and high data availability.