K-匿名隐私保护模型下的Top-k查询  

Top-k Queries under K-Anonymity Privacy Protection Model

作　　者：辛婷婷[1] 刘国华[1] 

机构地区：[1]东华大学计算机科学与技术学院,上海201620

出　　处：《计算机科学与探索》2011年第8期751-759,共9页Journal of Frontiers of Computer Science and Technology

基　　金：国家自然科学基金No.61070032~~

摘　　要：数据查询问题是K-匿名隐私保护模型下数据可用性问题之一。提出一种K-匿名数据的空间数据组织方法及其索引方法;定义了两种新的查询UK-Rank和NT-Rank,UK-Rank主要应用于一些需要排序的查询,NT-Rank应用于点查询或者范围查询;采用了Monte-Carlo积分近似计算的抽样方法来提高查询效率。对提出的相关算法进行了实验,结果表明,将K-匿名数据组织成空间数据的方法是可行的,并且应用抽样方法后,查询效率大大提高。How to answer queries under the K-anonymity privacy protection model is one of problems for the availability of anonymized data. This paper proposes a translation method from K-anonymized data to spatial data and an indexing method. It defines two new queries for the availability of anonymized data, UK-Rank and NT-Rank. UK-Rank is mainly used in queries that require sorting, NT-Rank is used in the point query or range query. The Monte-Carlo integration is used to compute accurate estimate of probability and improves query efficiency. Finally, related experiments are conducted. The experimental results show that the translation from K-anonymized data to spatial data is feasible, and the query efficiency is greatly improved after the application of sampling methods.

关 键 词：TOP-K查询 K-匿名数据 不确定数据库 偏序 R-TREE 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]