Complex systems simulation online visual analysis and assessment using dynamic aggregation operators

Agent-Based Simulations generate massive loads of data that need to be analyzed and visualized online. This paper proposes an approach that represents and abstracts dynamical properties of agent-based model using visual abstract operator. We present how spatio-temporal dynamic aggregation operators can facilitate online analysis tool of agent-based simulation. Spatial aggregation is used to represent multi level model with hierarchical and density-based clustering. Temporal aggregation is used for agent trajectory analysis and dynamic network analysis. Such an approach is related to data stream mining which is the process of extracting knowledge structures from continuous data records. It goes beyond visualization as the information displayed is the result of mining algorithms that are performed on the stream of data from the ABM simulation.