Agent Based Modeling of Moving Point Objects in Geospatial Data

We describe a framework for agent based modeling of moving point objects. Spatial movements are generated based on two overlapping ontologies. The first ontology includes the landmarks and descriptive outdoor behavior attributes. The second ontology includes indoor space and descriptive indoor behavior attributes. The modeling is based on ontology that includes the landmarks and descriptive behavior attributes of moving objects. The goal of our research in this area is to generate various spatial movements of point objects that can be classified into different known patterns. The agent behaviors can be modified semi-automatically based on changes in ontologies. This modeling is accomplished over a representation of a real-world data like road networks.