WIT Press


GIS/data Mining Applied For Identification Of Environmental Risk Factors For Diseases

Price

Free (open access)

Paper DOI

10.2495/MIS040061

Volume

32

Pages

6

Published

2004

Size

291 kb

Author(s)

S. Anno

Abstract

This study has tried to identify environmental risk factors for respiratory system diseases in Tokyo metropolitan area using GIS and data mining. GIS were applied to the analysis of spatial relationships between the distribution of the diseases and SPM exposure levels and distances from the national roads as environmental factors. A CART model as a data mining tool was applied for assessment of habitat types where there was a potential risk of the incidence of the diseases with the databases obtained from GIS analysis while identifying environmental risk factor levels concerned with the incidence of the diseases. Another purpose of this study was to provide a comprehensive outline of the methods of data mining, which have wide applications in a GIS and other areas where spatial data are used, whilst indicating its strengths and weaknesses, how and when to apply the methods and their role when used. Keywords: geographic information systems (GIS), spatial analysis, data mining, classification and regression trees (CART), decision tree. 1 Introduction Human respiratory effects of environmental factors are well documented; numbers of epidemiological studies have examined affect of environmental factors on respiratory health [1, 2, 3, 4, 5]. However, those studies have not analyzed spatial associations between respiratory system diseases and environmental exposures; moreover, those studies have not identified environmental risk factor levels concerned with the incidence of the diseases. Then, our study tried to analyze relationships between spatial pattern of the diseases and environmental factors; and identify environmental risk factor levels concerned with the incidence of the diseases using spatial analysis in geographic

Keywords

geographic information systems (GIS), spatial analysis, data mining, classification and regression trees (CART), decision tree.