文章摘要
张君维,杨静,张健沛,张乐君.面向数据流的敏感规则k 匿名保护算法[J].高技术通讯(中文),2013,23(3):251~256
面向数据流的敏感规则k 匿名保护算法
A k anonymity preservation algorithm for sensitive rules in data stream
  
DOI:
中文关键词: k 匿名, 数据流, 关联规则, 敏感规则
英文关键词: k anonymous, data stream, association rule, sensitive rule
基金项目:国家自然科学基金(61073041,61073043),黑龙江省自然科学基金(F200901),高等学校博士学科点基金(20112304110011)和优秀学科带头人专项资金(2011RFXXG015, 2010RFXXG002)资助项目
作者单位
张君维 哈尔滨工程大学计算机科学与技术学院 
杨静 哈尔滨工程大学计算机科学与技术学院 
张健沛 哈尔滨工程大学计算机科学与技术学院 
张乐君 哈尔滨工程大学计算机科学与技术学院 
摘要点击次数: 3252
全文下载次数: 2441
中文摘要:
      为了使攻击者通过降低阈值来发现被隐藏规则的概率小于1/k,以此实现对敏感规则的二重保护,提出了一种面向数据流的敏感规则k 匿名保护算法。该算法首先用时间滑动窗口技术来获取数据使用者最感兴趣的最新n个时刻到达的数据,然后通过追加交易的方式而并非修改交易项的方式来实现对敏感规则的k 匿名,从而避免对数据流的二次访问以及被隐藏敏感规则的泄漏。同时采用素数编码的方法,用素数集代替原始项集参与运算,以提高算法的效率和降低算法的空间复杂度。实验结果表明,此算法可以在数据流环境下高效进行敏感规则的k 匿名,并且能够保证挖掘结果的有用性。
英文摘要:
      To achieve the two fold protection for sensitive rules by lowering attackers’ probability of finding hidden rules to less than 1/k through reducing the thresholds of the minimum support threshold (MST) and the minimum confidence threshold (MCT), the paper proposes a data stream oriented k anonymity preservation algorithm for sensitive rules. The algorithm uses the time sliding window technique to get the latest n moment data which most interest data users, and then, completes k anonymity of sensitive rules by adding transactions data instead of modifying items of transactions, so it can avoid the twice access to data stream and the leaking of hidden sensitive rules. At the same time, the algorithm adopts prime number coding to further improve the efficiency and reduce the space complexity. The experimental results show that the proposed algorithm can achieve k anonymous sensitive rules hiding efficiently in data stream and keep the higher usefulness of data mining results.
查看全文   查看/发表评论  下载PDF阅读器
关闭

分享按钮