Fahem Kebair ABSTRACT The cirsis management is a complex problem raised by the scientific community currently. Decision support systems are a suitable solution for such issues, they are indeed able to help emergency managers to prevent and to manage crisis in emergency situations. However, they should be enough flexible and adaptive in order to be reliable to solve complex problems that are plunged in dynamic and unpredictable environments. The approach we propose in this paper addresses this challenge. We expose here a modelling of information for an emergency environment and an architecture of a multiagent decision support system that deals with these information in order to prevent the occur of a crisis or to manage it in emergency situations. We focus on the first level of the system mechanism which intends to perceive and to reflect the evolution of the current situation. The general approach and experimentations are provided here. INTRODUCTION Natural and man made disasters are permanent hazards for human beings since they may have harmful consequences for them and their properties. In order to brace such events, people must be efficient in their evaluations, their decision making and their actions.

The use of Decision Support Systems (DSSs) has considerably increased, during the last decade, due to the complexity of the problems faced by the decision makers. Indeed, the need for decision support tools should be, if anything, increasing [10]. In some domains or circumstances, making a decision is an arduous task that requires some abilities exceeding the human capacities. We can think decision-making in Simon's decision making model, which consists in intelligence, design and choice [11]. Based on this model, the complexity of decision making lies in the difficulty to get a clear insight into the problem to resolve, to process the vast amount of collected information, to make the right choice in time and to harmonise finally the set of decisions made by the decision makers or the organisations. Therefore, computer-based systems may be very helpful to support decision making, especially when the environment problem is complex, dynamic and partially known.

It is therefore difficult to actors to make good decisions in time and to coordinate efficiently their efforts, since they do not have enough knowledge about the situation or they do not have timely information they need. The emergency response is one of the greatest challenges that arise to the society currently. One approach to address this challenge is to develop decision support systems (DSS) that may help improve emergency planners and responders awareness and their decision-making abilities. Moreover the system must anticipate the risk of calamitous events or the evolution of a current crisis. This makes planners warned and prepared permanently to future events. Consequently, they can produce robust plans towards both short-term and long-term goals.

Making a decision in a changeable and dynamic environment is an arduous task owing to the lack of information, their uncertainties and the unawareness of planners about the future evolution of incidents. The use of a decision support system is an efficient solution of this issue. Such a system can help emergency planners and responders to detect possible emergencies, as well as to suggest and evaluate possible courses of action to deal with the emergency. We are interested in our work to the modeling of a monitoring preventive and emergency management system, wherein we stress the generic aspect. In this paper we propose an agent-based architecture of this system and we describe a first step of our approach which is the modeling of information and their representation using a multiagent system.

The topic of risk prevention and emergency response has become a key social and political concern. One approach to address thi s challenge is to develop Decision Support Systems (DSS) that can help emergency planners and responders to detect emergencies, as well as to suggest possible course of actions to deal with the emergency. Our research work comes in this framework and aims to develop a DSS that must be generic as much as possible and independent from the case study.

Recent catastrophic disasters have brought urgent needs for diverse technologies for disaster relief. Currently, there is an overwhelming need for better information technology to help support the efficient and the effective management of the disaster management (also known as emergency response). In particular, actors and agencies need an assistance to help them to make a decision in a fashion time and to be able to coordinate their efforts in a flexible way in order to prevent further problems or effectively manage the aftermath of a disaster. Our project is situated in this context and consists to develop a generic Decision Support System (DSS), able to detect a risk in an uncertain and partially perceived environment and to prevent its evolution. The DSS kernel is a multiagent system with three layers, where each one has a specific role. The role of the lower layer, that we call the representation layer, is to represent the environment state and its evolution over the time. The environment is perceived as a whole of entities, directly or indirectly observable and of which states change permanently. These entities are modeled according to a taxonomic organisa-Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes, University of Le Havre, 25 rue Philippe Lebon, 76058, Le Havre Cedex, France.

In an emergency situation, the actors need an assistance allowing them to react swiftly and efficiently. In this prospect, we present in this paper a decision support system that aims to prepare actors in a crisis situation thanks to a decision-making support. The global architecture of this system is presented in the first part. Then we focus on a part of this system which is designed to represent the information of the current situation. This part is composed of a multiagent system that is made of factual agents. Each agent carries a semantic feature and aims to represent a partial part of a situation. The agents develop thanks to their interactions by comparing their semantic features using proximity measures and according to specific ontologies.