时空轨迹相似性度量方法综述

Overview of the Similarity Measurement Methods for Spatial-temporal Trajectory

时空轨迹数据是记录移动对象时间和空间的位置序列,它是研究移动对象最为重要的数据来源。时空轨迹数据的分析挖掘是空间数据挖掘的一个研究热点,它包括轨迹检索、轨迹分类、模式挖掘、异常检测等。在分析挖掘过程中,轨迹之间的相似性度量是一个关键问题。本文研究时空轨迹相似性度量方法,首先从理论意义和应用价值的角度分析时空轨迹相似性度量的重要性,然后根据度量方式的不同将时空轨迹相似性度量方法分为两大类:基于轨迹点的相似性度量方法和基于轨迹段的相似性度量方法。由于基于轨迹点的各种相似性度量方法的应用场景和对相似性的定义不同,再将其细分为全局匹配度量法和局部匹配度量法。对时空轨迹相似性度量方法进行分类的同时,也对各个类别中常用的相似性度量方法进行了详细阐述,分析它们的优缺点及应用场合,为时空轨迹分析挖掘提供参考。

Spatial temporal trajectory data is one of the most important data sources of the research of moving object and it records the time and location of moving object.Spatial temporal trajectory data mining is a research focus of spatial data mining,which includes trajectory retrieval,trajectory classification,pattern mining,and anomaly detection and etc.The similarity trajectory measurement is one typical research point in the process of data analysis and mining.This paper studies the similarity measurement of spatial temporal trajectories.First,the importance of similarity measurement of spatial-temporal trajectories is analyzed from theoretical significance and practical value.Second,the methods of spatial temporal trajectory similarity measurement are divided into two categories based on the different ways of measurements： based on the trajectories points and the trajectory segment.As the application scenarios and the requirements for similarity of the various similarity measurements are variable,the methods based on the trajectories points are subdivided into overall matching method and partial matching method.This paper not only points out a reasonable classification of the spatial-temporal trajectory similarity measurement method,but also elaborates the similarity measure method commonly used in each category.Also,the advantages and disadvantages of each measurement method are analyzed,to provide a reference for spatial-temporal trajectory analysis.