Visual exploration of big spatio-temporal movement data

We propose a methodology for exploring groups of human objects and analyzing the movement behaviors of the groups within a specific given space. Movement datasets are generally too big to analyze. We have to use aggravation methods to reduce the datasets into groups. After exploring groups, we obtain an integrated solution with three visualization modules: Collision Detection View for visualizing the detailed information of each group, Force Direct Map for the community structure, and Hotspot View for finding which location in the given space is the most popular. Finally, we use millions of movement records to evaluate our method.