APP下载

一种基于GDK-means的高校贫困生信息隐私保护方法研究

2023-08-01刘晓娜王恺王成德徐彦强

现代信息科技 2023年11期

刘晓娜 王恺 王成德 徐彦强

摘  要:高校在对贫困生的资助过程中,为保证公开、公平,会获取相关学生很多关键性隐私数据,如贫困原因、生源所在地、家庭收入、家庭成员、在校消费等敏感隐私数据。同时资助的结果又要求必须公开以保证管理过程的公正性。针对高校对贫困生数据发布中的公开与隐私保护之间的矛盾,提出了一种基于GDK-means的隐私保护方法。在该算法下,在K-means聚类的基础上,对生成的簇进行簇内泛化,来对发布的敏感数据进行去隐私化处理,以达到用户隐私保护的目的,同时量化了处理所带来的信息丢失度。经理论分析和实验,验证了采用GDK-means算法,在保证数据可用性的前提下,可实现数据发布中较好的隐私保护性。

关键词:贫困生资助;分布式聚类;隐私保护;K-means算法

中图分类号:TP311.13;G647  文献标识码:A  文章编号:2096-4706(2023)11-0030-04

Research on a College Information Privacy-Protection Method for Poor Students

Based on GDK-means

LIU Xiaona1, WANG Kai1, WANG Chengde1, XU Yanqiang2

(1.Lanzhou University of Arts and Science, Lanzhou  730000, China; 2.Lanzhou Institute of Technology, Lanzhou  730050, China)

Abstract: In the process of providing financial aid to poor students, colleges and universities obtain a lot of key private data about the students in order to ensure openness and fairness, such as the reason for poverty, the location of the student's origin, family income, family members, school spending and other sensitive private data. At the same time, the results of financial aid must be made public to ensure the fairness of the management process. A privacy-protection method based on GDK-means is proposed to address the contradiction between disclosure and privacy-protection in the release of data on poor students in colleges and universities. Under this algorithm, the published sensitive data can be de-privatised by intra-cluster generalisation of the generated clusters on the basis of K-means clustering to achieve the purpose of user privacy-protection, while quantifying the degree of information loss caused by the processing. After theoretical analysis and experiments, it is verified that the use of GDK-means algorithm can achieve better privacy-protection in data publishing under the premise of ensuring data availability.

Keywords: financial aid for poor students; distributed clustering; privacy-protection; K-means algorithm

0  引  言

高校在對生活困难学生认定时,主要标准就是家庭条件困难和在校期间生活简朴。其中评定时需要收集学生在校日常消费数据,以及影响家庭经济状况的有关因素开展认定工作,如家庭收入、家庭负担、特殊群体、生源地、突发状况、学生食堂和教育超市消费等数据,以供对学生进行资助和相关等级认定,并在完成评定后需要对学生的部分信息进行公示,才能保证整个评定过程公平、可靠。但是,这些原始数据中通常包含很多的个人隐私信息,如果不经过任何处理就直接发布,势必会造成严重的隐私泄露,从而导致困难身份披露和属性披露,使得善意的助学行为变成对困难学生的心理伤害,甚至会降低部分困难学生的参与度,从而降低了资助的善意效力。……

登录APP查看全文