基于多维泛化路径的K-匿名算法  

K-anonymity Algorithms Based on Multi-Dimensional Generalization Path

作　　者：黄春梅[1] 费耀平[1] 李敏[1] 戴弋[1] 刘娇[1] 

机构地区：[1]中南大学信息科学与工程学院,长沙410075

出　　处：《计算机工程》2009年第2期154-156,共3页Computer Engineering

摘　　要：为使微数据发布在满足K-匿名要求的同时提高匿名数据的精度,提出多维泛化路径的概念及相应的2种K-匿名算法,包括完整FilterK-匿名算法和部分FilterK-匿名算法。将它们与Incognito算法和Datafly算法进行比较,实验结果表明2种算法都能有效降低匿名信息损失,提高匿名数据精度和处理效率。Microdata publication need satisfy the basic K-anonymity requirement as well as improve the precision of anonymized data. This paper proposes two related K-anonymity algorithms based on the notion of multi-dimensional generalization path, namely K-anonymity Filter algorithm and K-anonymity partial Filter algorithm. In comparison with classic Datafly algorithm and Incognito algorithm, the two algorithms offer more efficiency for both reducing anonymization cost and improving data precision.

关 键 词：K-匿名 微数据 隐私保护 域泛化层次结构 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]