Narrative Visualisation of Simulations: Finding the Stories Within the Data

Abstract

Decision makers need to understand and draw key insights from large and complex data sets generated by simulations. This may be to answer a specified analytical problem addressed by a simulation or to support learning or mission objectives through after-action review of a virtual simulation. Geospatial displays are traditionally used to help a decision maker or analyst gain an understanding of the results of a force-on-force constructive simulation, or for after-action review of virtual simulations. They provide an evolving picture of the geospatial positions and spatial relationships between entities and key features in the environment, and can potentially incorporate other attributes such as entity state or key metrics to support the analyst. However, while this geospatial context is often important, sense-making also requires an understanding of the significant actors, their relationships, their interactions, and the consequences of these relationships and consequences. Narrative visualisation is an approach that focusses on the storylines of key actors, the events they are involved in, and the causal relationships between entities-i.e. the stories within the data. This paper describes initial work demonstrating a novel narrative visualisation approach for data from a large-scale force-on-force simulation, and the key features of this visualisation that supports a decision maker or analyst's understanding and insights on what occurred within the simulation, and why. To the authors' knowledge, this represents the first time simulation data has been applied to narrative visualisation. While still in its early stages, this approach shows promise and could potentially supplement, or indeed replace, geospatial-centric visualisations as the primary sense-making pathway for analysts in a wide range of operating domains, including real-time operating pictures. This is particularly apropos in domains where the geospatial context is less relevant.