Rough Set Analysis to Manage Urban Air Pollution Control System

The expansion of urban areas causes many negative impacts both in urban areas and in the surrounding countryside. Scattered urban structures increase energy, land and soil consumption, they raise greenhouse gas emissions that cause climate change and increase traffic and consequently also air and noise pollution. Pollution in urbanized areas strongly depends on the degree of urbanization and on the meteorological conditions; the main contributions of pollutants come from sources of combustion, above all from means of transport and urban heating. This work shows the application of an innovative technique of data mining, the Rough Set Analysis, to the study of two-years carbon monoxide (CO) concentrations in Catania town (Italy). The study aims to investigate about parameters locating more influential CO dispersion phenomena in order to evaluate the policies of intervention to reduce the urban air pollution levels. This methodology explains the above mentioned relationships in terms of “if…, then…” sentences and can be considered as an efficient decision support system which aims at the improvement of the urban air quality management. In order to individuate the relation between condition and decision criteria expressed in the form of decision rules, preliminarily were calculated the condition – decision correlations. Finally, the decision criterion was discretised in 4 decision classes, according to the Italian law, in order to evaluate the daily effects after 24 hours by the application of the decision rules in terms of urban air pollution; the important results are presented by some tables with the most significant minimal decision rules.