时空轨迹相似性度量方法探讨

Discussion on Similarity Measurement Method of Spatiotemporal Trajectory

时空轨迹的相似度计算将轨迹间相似特征进行了量化,是上游相似查询和分析应用的基础,有效而稳定的相似性度量及其高效计算方法也是众多轨迹模式挖掘的必要前提。本文研究时空轨迹相似性度量方法,首先探讨了时空数据模型和轨道数据模型的研究现状;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,即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。

The similarity calculation of spatiotemporal trajectories quantifies the similarity characteristics between trajectories,which is the basis of upstream similarity query and analysis application.Effective and stable similarity measurement and efficient calculation method are also the necessary premise of many trajectory pattern mining.In this paper,it studies the spatiotemporal trajectory similarity measurement methods.Firstly,the research status of spatiotemporal data model and trajectory data model are discussed.Then,according to the different measurement methods,the spatiotemporal trajectory similarity measurement methods are divided into two categories:the similarity measurement method based on trajectory points and the similarity measurement method based on trajectory segments.Due to the different application scenarios and definitions of the similarity measurement methods based on the trajectory points,they are subdivided into global matching measurement method and local matching measurement method.