基于MapReduce的并行抽样路径K-匿名隐私保护算法

A parallel sampling path K-anonymity privacy protection based on MapReduce

K-匿名算法及现存K-匿名改进算法大多使用牺牲时间效率降低发布数据信息损失量的方法实现数据的匿名化,但随着数据量的急剧增长,传统的数据匿名化方法已不适用于对较大数据的处理。针对K-匿名算法在单机执行过程中产生大量频繁项集和重复搜索数据表的缺点,将MapReduce模型引入到抽样泛化路径K-匿名算法中对其进行优化。该方法兼具MapReduce及抽样泛化算法的优点,高效分布式匿名化数据集,降低发布数据集信息损失量,提高数据的可用性。实验结果表明:当数据量较大时,该优化算法在时间效率及数据精度方面有显著提高。

K-anonymous algorithm and improved algorithm K-anonymous mostly use the method of sacrificing of time to lower data information loss to realize the data anonymity, but with the rapid growth of data quantity, the traditional methods of data anonymity is not suitable for processing of large data. Aimed at the shortage of time complexity and execution efficiency K-anonymity in stand-alone that it generates a lot of frequent sets and searches the data table repeatedly, this paper introduces the MapReduce technology to K-anonymity algorithm to optimize this algorithm. The algorithm with the advantage of MapReduce and sampling generic algorithm can compute distributed anonymous data set effectively and reduce the information loss of released data set, so it improves the availability of data. The experimental results show that the algorithm increases observably in time efficiency and data accuracy.