Within the „Forensic Disaster Analysis“ various institutes of KIT and the Geoforschungszentrum Potsdam are working together on an interdisciplinary basis. The knowledge database to be developed in this subproject has a cross-sectoral task since several subprojects intend to use it. The knowledge database is a combination of existing and event-specific databases containing historical disasters such as earthquakes, floods, or extreme weather events. This allows further usage of analysis tools in the frame of CEDIM FDA. In particular, we focus on case-based reasoning which is a methodology that utilizes similar past events in order to draw conclusions about a new and to a large extent unknown event. The historical database can either be extended by experts or new cases over the course of time.
The first step towards a structured storage of historical data is the data modeling in particular the definition of attributes for the event description. Therefore we have to analyze how to compare two events and which attributes are relevant in this context. Moreover the required parameters containing information from past events that are most relevant for the current analysis have to be defined. The structure of the database will be flexible allowing different kinds of disasters to be covered. Additionally, a mathematical similarity function will be developed in order to retrieve similar events from historic data for using them in the new analysis. The generic approach of the knowledge database and the case-based reasoning application allows the shared use by all institutes working in the project.
Within the first year of the project the basic structure of the knowledge database and the similarity function will be defined. Subsequently the database will be developed and the retrieval step will be implemented. In the following years the algorithms will be enhanced and the database structure will be extended allowing further disaster types to be dealt with.
Pic. 1: The CEDIM database is a combination of existing databases and the methodology case-based reasoning provides an analysis support for other participating project partners. The CBR cycle illustrated is based on Aamodt and Plaza, Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches, 1994.
S. Moehrle, “Generic self-learning decision support system for large-scale disasters,” in ISCRAM 2012 Conference Proceedings - 9th International Conference on Information Systems for Crisis Response and Management, 2012.
S. Moehrle, “Modeling of countermeasures for large-scale disasters using high-level petri nets,” in ISCRAM 2013 Conference Proceedings - 10th International Conference on Information Systems for Crisis Response and Management, 2013.
S. Moehrle, “Towards a decision support system for disaster management,” in Safety, Reliability and Risk Analysis: Beyond the Horizon - Proceedings of the European Safety and Reliability Conference, ESREL 2013, 2013, pp. 239–246.
B. Schröter, K.; Mühr, B.; Kunz, M.; Elmer, F.; Khazai, B.; Bessel, T.; Möhrle, S.; Dittrich, A.; Kunz-Plapp, T.; Kreibich, H.; Trieselmann, W.; Merz, “Das Hochwasser im Juni 2013 in Deutschland: Auslöser, Ausprägung und Auswirkungen,” Wasser - Landschaft - Mensch in Vergangenheit, Gegenwart und Zukunft, Forum für Hydrologie und Wasserbewirtschaftung 34.14, 2014.
S. Möhrle and W. Raskob, “Case-based analytical support for rapidly assessing natural disasters,” in 14. Forum Katastrophenvorsorge, Leipzig, 2014.
K. Schröter, F. Elmer, W. Trieselmann, H. Kreibich, M. Kunz, and B. Khazai, “The near real time Forensic Disaster Analysis of the central European flood in June 2013 - A graphical representation of the main results,” in Geophysical Research Abstracts, EGU2014-6054, General Assembly European Geosciences Union 2014, 2014, vol. 16.
B. Khazai, A. Benkler, T. Bessel, S. Möhrle, and K. Schröter, “Analysis of the Potential Resilience and Preparedness of Districts affected by the June 2013 Flood in Germany,” in Geophysical Research Abstracts, EGU2014-6054, General Assembly European Geosciences Union 2014, 2014, vol. 16.
B. Khazai, T. Bessel, S. Möhrle, A. Dittrich, and K. Schröter, “Near real time Forensic Disaster Analysis of the central European flood in June 2013 in Germany : Impact and management,” in Geophysical Research Abstracts, EGU2014-6054, General Assembly European Geosciences Union 2014, 2014, vol. 16.
S. Moehrle, “On the assessment of disaster management strategies,” in ISCRAM 2014 Conference Proceedings - 11th International Conference on Information Systems for Crisis Response and Management, 2014.
S. Möhrle and B. Mühr, “Case-based damage assessment of storm events in near real-time,” in Geophysical Research Abstracts, EGU2015-12293, 2015 EGU General Assembly 2015, 2015, vol. 17.