Understanding the impact of temporal scale on human movement analytics
Understanding the impact of temporal scale on human movement analytics
复制标题
DOI：
10.1007/s10109-021-00370-6
复制
发表时间：
2022-02
影响因子：
2.9
通讯作者：
Rongxiang Su;S. Dodge;K. Goulias
中科院分区：
文献类型：
--
作者：
Rongxiang Su;S. Dodge;K. Goulias
运动是通过一系列模式在多个空间和时间尺度上表现出来的。今天的运动数据变得越来越细粒度的时间粒度。这些观测结果往往代表了运动路径上的多种行为模式和沿着的复杂模式，然而，运动数据的观测尺度与运动模式捕获的分析尺度之间的关系仍然研究不足。本文旨在研究时间尺度在运动数据分析中的作用。它提出了一个重要的问题，即“围绕运动数据和分析的规模的决策如何影响我们对运动模式的推断？”通过在人类运动背景下的一组计算实验，我们系统地研究了不同时间尺度对常见运动分析技术的影响，包括轨迹分析以计算运动参数（例如，速度、路径曲折度）、个体空间使用的估计以及用于检测多个移动的个体之间的潜在接触的交互分析。
Movement is manifested through a series of patterns at multiple spatial and temporal scales. Movement data today are becoming available at increasingly fine-grained temporal granularity. These observations often represent multiple behavioral modes and complex patterns along the movement path. However, the relationships between the observation scale of movement data and the analysis scales at which movement patterns are captured remain understudied. This article aims at investigating the role of temporal scale in movement data analytics. It takes up an important question of “how do decisions surrounding the scale of movement data and analyses impact our inferences about movement patterns?” Through a set of computational experiments in the context of human movement, we take a systematic look at the impact of varying temporal scales on common movement analytics techniques including trajectory analytics to calculate movement parameters (e.g., speed, path tortuosity), estimation of individual space usage, and interactions analysis to detect potential contacts between multiple mobile individuals.