近年来由于移动设备上全球定位(GPS)功能的普及，社会化媒体上传播的信息不仅含有时间属性，还往往附带有空间属性，因此为研究信息的时空传播提供了难得的契机。尽管如此，快速地观察和分析信息的时空传播，尤其是大尺度(例如全国乃至全球)传播仍然异常困难，原因主要有：1)时空传播规律隐藏在大规模异构的复杂用户行为数据中，难以进行直接研究；2)缺乏有效的针对时空传播行为的时空数据可视化方法和技术。..本项目研究基于时空维度的社会化媒体用户传播信息行为的可视分析方法，解决三个主要问题：1)建立可靠和可用的时空传播定量模型，迅速发现并定量评估多个话题在社交网络上的空传播的强度和途径；2)设计直观和有效的时空可视化方法，展现多层次、多维度和多主题的时空传播过程；3)开发易用和合适的时空可视化交互技术，构建时空可视分析系统，允许用户实时地监控、分析和追踪信息时空传播的轨迹。

The advance of technology leads to the prevalence of Global Positioning System (GPS) in mobile devices, which help produce a huge volume of information spreading on social media that carries not only temporal information but also geographic information. This provides a great opportunity to detect and study the patterns of spatio-temporal diffusion of information on social media. Nevertheless, quick detection and analysis of the spatio-temporal diffusion of information, especially in a large (such as global) scale is rather challenging. First, interesting spatio-temporal diffusion (i.e., communication among various groups of users in different locations over time) patterns are usually hidden in large-scale, complex user behavior data, which is difficult to observe and analyze directly. Second, effective and intuitive visualization of spatio-temporal diffusion of information is still absent in the field. ..This project will focus on visual analytics of spatio-temporal diffusion of information on social media and study the following problems: 1) To derive a reliable and quantitative model that can quickly detect interesting spatio-temporal diffusion patterns from the large-scale, complex user behavior data and quantitatively assess the strength of the diffusion; 2) To design intuitive and effective visualization methods to help analysts explore and understand the interesting spatio-temporal diffusion patterns detected by the quantitative model; 3) To develop easy-to-use and proper user interaction techniques for spatio-temporal visualization, and build a visual analytics system that enables analysts to monitor, analyze, and track the diffusion of information in real time.