Fuzzy spatial association rule mining to analyze the effect of environmental variables on the risk of allergic asthma prevalence

Abstract

The prevalence of allergic diseases has greatly increased in recent decades, likely due to contamination of the environment with allergy irritants. One common treatment is identifying that allergy irritant, and then avoiding exposure to it. This article studies the relation between the prevalence of allergic asthma and certain allergy irritants that are related to environmental variables. To that end, we use spatial association rule mining to determine the association between the spatial distribution of allergic asthma prevalence and air pollutants such as CO, SO2, NO2, PM10, PM2.5, and O3 (from data compiled by air pollution monitoring stations), as well as other factors, such as the distance of residence from parks and roads. In order to clear up the uncertainties inherent in the attributes linked to the spatial data, the dimensions in question have been defined as fuzzy sets. Results for the case study (i.e. Tehran metropolitan area) indicate that distance to parks and roads, as well as CO, NO2, PM10, and PM2.5 levels are related to allergic asthma prevalence, while SO2 and O3 are not. Finally, we use the extracted association rules in fuzzy inference system to produce the spatial risk map of allergic asthma prevalence, which shows how much is the risk of allergic asthma prevalence at each point of the city.

Keyword : spatial data mining, fuzzy spatial association rule mining, risk analysis, allergic asthma, air pollution, GIS