Agent-Based Decision Making through Intelligent Knowledge Discovery Marina V. Sokolova 1,2 andAntonioFern´andez-Caballero 1 1 University of Castilla-La Mancha, Polytechnical Superior School of Albacete, Campus Universitario s/n, 02071-Albacete, Spain smv1999@mail.ru,caballer@dsi.uclm.es 2 Kursk State Technical University, Kursk, ul.50 Let Oktyabrya, 305040, Russia Abstract. Monitoring of negative effects of urban pollution and real- time decision making allow to clarify consequences upon human health. Large amounts of raw data information describe this situation, and to get knowledge from it, we apply intelligent agents. Further modelling and simulation gives the new knowledge about the tendencies of situa- tion development and about its structure. Agent-based decision support system can help to foresee possible ways of situation development and contribute to effective management and planning. 1 Introduction Humans, as the active element of environment, interact with it and receive the feedback, that is not always positive though, and sometimes appears to be un- healthy or even harmful. Negative environmental impact upon human health is a complicated one and can be, summed up from permanent effects, such as sun radiation, Earth´s geomagnetic field; from human´s life and work conditions (characteristics of profession, quality of portable water, air, ) which also include characteristics of the place they live in - transport and industrial activity, use of vehicles; and from instable characteristics like metheorological conditions. Humans adapt to natural unhealthy factors, often, at the cost of the organ- ism´s reversible and irreversible disfanctions, which can be then inherited and came to light as birth misfunctions and defects. During a long time an expert-system paradigm remained a standard solution for environmental assessment issues, as this approach supplied researches with all the necessary instrumentation: sets of expert information and analilical sup- port data. Later, appeared multi-agent system based applications, called by the requirements to real-time decision making and intelligent data mining. Some authors apply agent-approach to emergency assessment system devel- opment [1] related to traffic assidents responce by allowing fast situation de- scription, generation possible solutions and selection the decision maker for that case. MAS approach was utilyzed to increase the efficiency of hospital processes and to foresee possible effects of changes that can happen when expenses are reduced and the health-care budgets are restricted and stabilyzed. Different re- search institutes and organizations confirm the ecological contribution to illness