面向隐私保护的分布式数据挖掘关键问题研究

项目名称： 面向隐私保护的分布式数据挖掘关键问题研究

项目编号： No.61075053

项目类型： 面上项目

立项/批准年度： 2011

项目学科： 轻工业、手工业

项目作者： 申艳光

作者单位： 河北工程大学

项目金额： 10万元

中文摘要： 本课题对面向隐私保护的分布式数据挖掘关键问题进行深入研究，完成了原定研究计划并取得了以下主要成果：(1)建立了多维敏感属性下的多粒度隐私保护模型，给出了相应的隐私保护决策度定义、粒化匿名算法和个性粒化k-匿名的泛化匿名算法；(2)提出了一种面向隐私保护的个性化搜索结果排序方法；(3)提出了一种两层传感器网络中可验证隐私保护 查询协议SafeTQ（Safe Query）；(4)提出了基于项目聚类的改进的协同滤波算法、分布式隐私保护的聚类算法、隐私保护的多层关联规则挖掘算法和基于服务访问日志的服务发现方法。基于上述工作，本项目共发表论文14 篇，其中国际刊物论文3 篇、国际会议论文2 篇，已被EI检索1篇。培养研究生多名。我们所做的工作能够促进面向隐私保护的分布式数据挖掘的研究及应用。

中文关键词： 数据挖掘；隐私保护；安全多方计算

英文摘要： This project has studied the key problems of privacy-preserving distributed data mining, and has accomplished the research plan and obtained the following results: (1) proposed a new k-anonymity model of privacy preserving, the corresponding definitions of the individual privacy-preserving decision degree, the algorithms of granularity anonymity and the generalization anonymous algorithm of the anonymity were also given; (2) proposed a ranking method of personalized search result for privacy preservation; (3) proposed a verifiable privacy-preserving top-k query protocol in two-tiered sensor networks (SafeTQ); (4)proposed an improved item-based collaborative filtering algorithm, a privacy preserving distributed clustering algorithm, a privacy preserving multi-level association mining, a services discovery method based on service access log. Based on the above results of this project,14 papers have been published, including 3 papers in international journals, 2 papers in international conferences. 1 papers have been indexed by EI. The project has also supported a number of graduate students. Our work can promoted the research and applications of privacy-preserving distributed data mining.

英文关键词： data mining; privacy preserving;secure multi-party computation