Visual analysis of movement and event data in spatiotemporal context

定位技术的进步使得能够收集关于不同领域中不同类型对象的运动的海量数据，这提出了对用于分析这些数据的可扩展方法的需求。作为回应，最近在数据挖掘、地理可视化、信息可视化和视觉分析方面出现了许多方法和工具。然而，这些方法中的大多数只处理移动数据，而没有考虑移动的时空上下文，包括不同地点和不同时间的属性以及影响和/或受移动影响的各种空间、时间和时空对象。该项目旨在开发理论基础和新的可扩展的方法来分析上下文中的运动，使用以数据集的形式获得的显性上下文信息以及人类分析员头脑中可用的隐含上下文信息。该方法将交互式视觉界面与计算技术相结合，以支持人机协同。该项目将开发能够从对移动数据的探索性分析过渡到创建表示视觉分析过程结果的显式正式模型的方法。在理论部分，该项目将开发一个移动的概念模型，它的时空背景，以及移动和背景之间可能的关系。在此基础上，该项目将建立分析任务分类和方法分类，为根据分析目标和要分析的数据的特征选择分析方法和工具提供指导方针。开发的理论和方法将通过原型软件工具的创建和在实际应用场景中的评估来验证。

Progress in positioning technologies has enabled the collection of huge amounts of data about movement of diverse types of objects in various domains, which has raised a demand for scalable methods for analyzing such data. In response, a number of methods and tools have appeared recently in data mining, geographic visualization, information visualization, and visual analytics. However, most of these approaches deal with movement data alone without taking into account the spatiotemporal context of the movement, which includes the properties of different places and different times and various spatial, temporal, and spatiotemporal objects affecting and/or being affected by the movement. This project aims at developing theoretical foundations and novel scalable methods for analyzing movement in context with the use of explicit context information available in the form of datasets as well as implicit context information available in the mind of human analyst. The methods will combine interactive visual interfaces with computational techniques for supporting synergistic collaboration of human and computer. The project will develop approaches enabling the transition from exploratory analysis of movement data to creation of explicit formal models representing the results of the visual analytics processes. In the theoretical part, the project will develop a conceptual model of movement, its spatiotemporal context, and possible relations between movement and context. On this basis, the project will build a taxonomy of analysis tasks and a taxonomy of methods, which will provide guidelines for choosing analytical methods and tools depending on the analysis goals and characteristics of the data to analyze. The developed theory and methodology will be verified through creation of prototype software tools and their evaluation in real application scenarios.