时空轨迹相似度计算方法研究与实现  

Research and Implementation of Similarity Computation for Spatiotemporal Trajectories

作　　者：涂刚凯 耿鑫 TU Gangkai;GENG Xin(Nanjing Fiber Home World Communication Technology Co.,Ltd.,Nanjing 210019;Wuhan Research Institute of Posts and Telecommunications,Wuhan 430074)

机构地区：[1]南京烽火天地通信科技有限公司,南京210019 [2]武汉邮电科学研究院,武汉430074

出　　处：《计算机与数字工程》2020年第5期1114-1120,共7页Computer & Digital Engineering

摘　　要：移动互联网时代的到来,产生了海量的时空轨迹数据,使得挖掘时空轨迹的潜在特征成为研究的热点,而人员时空轨迹是其中的一个方面,如何从海量的轨迹中挖掘出人员的关系人,涉及到优秀的判定算法以及高效的计算。针对传统的LCSS算法在轨迹点比对时,出现时间阈值选取的敏感性问题,提出了LCSS+算法,在不同的时间阈值条件下,表现平稳,识别率高。并针对大数据量的问题,提出了分布式环境下的LCSS+算法,实验结果表明分布式LCSS+算法能够缩短比对的时间,提升大数据集情形下的实时性。The arrival of mobile Internet era has generated massive spatiotemporal trajectory data.Making the potential characteristics of temporal and spatial trajectories a hot topic of research.People’s temporal and spatial trajectories is one aspect.How to mine the relationship person from the massive trajectory involves excellent decision algorithm and efficient calculation.In view of the sensitivity of the traditional LCSS algorithm to the selection of time threshold,a LCSS+algorithm is proposed.Under different time threshold conditions,the performance is stable and the recognition rate is high.In order to solve the problem of large amount of data,a LCSS+algorithm in distributed environment is proposed.The experimental results show that the distributed LCSS+algorithm can shorten the comparison time and improve the real-time performance of the case of large datasets.

关 键 词：时空轨迹 LCSS 伴随模式 HADOOP 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]