Geospatial Temporal Evidential Reasoning System (GEO-STEERS)

During Phase I of this research, we developed the GeoSpatial Temporal Evidential Reasoning System (GEOSTEERS) Phase I to III vision. GEOSTEERS provides the necessary capabilities for users to interactively transfer subject matter knowledge and experience into computational probability models. GEOSTEERS provides the following capabilities: discovery of relationships, patterns and connections between diverse sources of geospatial and/or temporal data, identification of important geospatial and temporal relationships and generation of hypotheses and predictive models significant event and continual improvement of our geospatially and temporally enabled predictions through an advanced data association technique. Some significant features of GEOSTEERS include: the capability to identify key geospatial entities (events, persons, locations, organizations or objects) from structured and unstructured data, the ability to quantitatively evaluate many combinations of relationships between suspect and non-suspect entities and events, a complete integration with ESRI’s ArcGIS for visualization of the relationship information and predictive models, a hypothesis management capability that can handle large, computationally complex relationships and interactions and a continuous searching capability that dynamically evaluates new possible relationships between events as new data is observed