基于DTW距离度量函数的DTW-TA轨迹匿名算法
发布时间：2018-10-08 18:32
【摘要】：在传统的基于欧几里德距离函数的轨迹相似性计算过程中,要求轨迹等长且时间点对应,无法度量不等长且有局部时间偏移的轨迹相似性。因此在构造同步轨迹集合过程中产生信息损失较大,影响轨迹数据的可用性。为此,通过引进一种可以度量不等长且有局部时间偏移的轨迹间相似性的DTW(dynamic time warping)距离度量函数,提出一种新的轨迹匿名模型——(k,δ,p)-匿名模型,构造了DTW-TA(dynamic time warping trajectory anonymity)算法。在合成数据集和真实数据集下的实验结果表明,该算法在满足轨迹k-匿名隐私保护的基础上,减少了信息损失,提高了轨迹数据的可用性。
[Abstract]:In the traditional trajectory similarity calculation based on Euclidean distance function, the trajectory similarity with equal length and time points is required, and the locus similarity with unequal length and local time offset can not be measured. Therefore, the information loss in the process of constructing synchronous track sets is large, which affects the usability of trajectory data. Therefore, by introducing a DTW (dynamic time warping) distance metric function which can measure the similarity between tracks with unequal length and local time offset, a new locus anonymous model- (k, 未 -p) -anonymity model is proposed, and the DTW-TA (dynamic time warping trajectory anonymity) algorithm is constructed. The experimental results on synthetic data sets and real data sets show that the proposed algorithm can reduce the loss of information and improve the usability of trajectory data on the basis of satisfying the path k- anonymity privacy protection.
【作者单位】： 江西理工大学信息工程学院;
【基金】：国家自然科学基金资助项目(61462034,61563019) 江西省教育厅科学技术研究项目(GJJ13415) 江西理工大学科研基金重点课题(NSFJ2014-K11)
【分类号】：TP309
，
本文编号：2257832
[Abstract]:In the traditional trajectory similarity calculation based on Euclidean distance function, the trajectory similarity with equal length and time points is required, and the locus similarity with unequal length and local time offset can not be measured. Therefore, the information loss in the process of constructing synchronous track sets is large, which affects the usability of trajectory data. Therefore, by introducing a DTW (dynamic time warping) distance metric function which can measure the similarity between tracks with unequal length and local time offset, a new locus anonymous model- (k, 未 -p) -anonymity model is proposed, and the DTW-TA (dynamic time warping trajectory anonymity) algorithm is constructed. The experimental results on synthetic data sets and real data sets show that the proposed algorithm can reduce the loss of information and improve the usability of trajectory data on the basis of satisfying the path k- anonymity privacy protection.
【作者单位】： 江西理工大学信息工程学院;
【基金】：国家自然科学基金资助项目(61462034,61563019) 江西省教育厅科学技术研究项目(GJJ13415) 江西理工大学科研基金重点课题(NSFJ2014-K11)
【分类号】：TP309
，
本文编号：2257832
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2257832.html
