机构地区: 北京师范大学信息科学与技术学院
出 处: 《广西师范大学学报(自然科学版)》 2009年第3期97-100,共4页
摘 要: 改进了局部线性嵌入LLE变换中的两个参数最近邻点数目K和高维空间维数d的选择方式,然后将数据集的敏感属性用随机LLE方法嵌入到高维空间中,从而使敏感属性得到加密保护。由于LLE方法可以保持数据集的内部拓扑结构,可使加密变换后的数据集以同样的精度找到正确的最近邻点。通过在一个具体旅游数据集上的实验结果表明,该算法能使安全性和精确度都达到要求。 A nearest neighbor tour circuit encryption algorithm based random LLE transformation is proposed. First alter the selection of the two parameters of the nearest neighbor numbers K and dimension d,then embed the tourists' sensitive attribution into higher dimension space using random LLE,so the sensitive attributes are encrypted and protected. Because LLE method can keep topology structure of dataset,the nearest points are kept with 100% precision on transformed dataset. The experiment shows that the present method can provide enough protect for tourists' sensitive information ,and it also gives the tourists appropriate tour circuit.
领 域: [自动化与计算机技术] [自动化与计算机技术]