【摘要】 对大数据查询过程机密数据低延时发布协议,能够高效解决机密数据查询时间过长的问题。对大数据查询过程中机密数据低延时发布协议的研究,需要对机密数据矩阵按特征差异分为不同数据簇,以簇中心变动的相对距离确定低延时值,完成机密数据低延时发布协议。传统方法构建一个机密数据低延时信息特征模型,对低延时属性进行快速定位,但忽略了确定低延时值,导致发布协议效果不理想。提出基于曼哈顿度量的大数据查询过程机密数据低延时发布协议方法,用信息增益法对机密数据矩阵分类,对分类后的矩阵进行机密数据特征提取,并按特征差异分为不同数据簇,采用曼哈顿度量算法来衡量数据簇中心的前后变化,以簇中心变动的相对距离确定低延时值,通过比较低延时值与所选择的低延时阈值大小来判断机密数据的低延时情况。仿真结果表明,上述方法发布协议效果较理想,效率更高,耗时更少。更多还原

【Abstract】 To research the low-delay publishing protocol of confidential data during big data query needs to divide the confidential data matrix into different data clusters based on feature difference. The traditional method builds a low-delay information feature model of confidential data, but ignores the low-delay value, leading to unsatisfactory result of publishing protocol. This paper puts forward a method of low-delay publishing protocol for confidential data during big data query based on Manhattan metrics. Firstly, this method used the information gain method to classify confidential data matrix and extracted confidential data feature from the classified matrix, and then divided them into different data clusters based on feature difference. Moreover, this research used Manhattan metric algorithm to measure the change of data cluster center, so as to determine the low-delay value based on the relative distance of change of cluster center. Finally, we judged the low-delay situation of confidential data by comparing the low-delay value with the selected low-delay threshold. Simulation results show that the proposed method has good effect and high efficiency of publishing protocol. Meanwhile, this method is less time-consuming.更多还原