X.Y. Liu1, H. Zhang1, G.N. Lu1, A.A. MacEachren2

1 - Provincial Key laboratory of GIScience, Department of Geography, Nanjing Normal University

2 - GeoVISTA Center, Department of Geography, Penn State University


In each factor of disaster (from preparedness, prevention, response, and eventual recovery), all require access to geographically referenced information and tools for making available information relevant to the tasks at hand and in time. Geographic Information and technologies (GI&T) are clearly and incontestably an essential part of disaster management in all aspects and have been used effectively at all scales. However, current GI&T specifically focus on Geospatial abstraction methods, more specifically on spatial data acquisition, storage, update, query, representation and visualization. This frequently leads to insufficient use of implicit knowledge and explicit knowledge as well due to the lack of awareness of participants. So the question of what information are quoted, and how to document, process and distribute valuable information to responders effectively, remains an open research topic in crisis management domain, and many other GIS application fields as well.

This paper proposes an approach to extend traditional cartographic visualization to cartographic process visualization by combining cognitive tools (ontology-based knowledge management - concept maps) in the process of map generation in the context of crisis management. Specifically, we focus on how the interactive maps can automatically be generated based on a particular or given scenario by building generating patterns and transferring mechanisms between visual representation of formal knowledge and Geo-database through concept maps. The potential of this approach is supposed to facilitate visual, manipulable, dynamic, and analytical cognitive process in geographic visualization, thus to enhance visual availability of complex phenomena or processes in a time strict situation, and to support decision-making effectively.

In this paper, Emergency Response of Environmental Pollution Event is designed as a characteristic scenario background. Scenario is logically structured as formal ontologies, graphically represented by concept maps. The scenario can be recomposed, and/or recombined to form new knowledge in terms of multiple viewpoints of users.

One of the main challenges in this research is how interactive maps can be generated automatically by bridging across concept maps. It can be addressed through encoding semantic mapping between concept maps and interactive maps to build the relationships of visual representation structures of particular problems and geographic visualization patterns.