Mapping urban function by integrating socio-economic information, address data, and remote sensing

The paper presents an approach for mapping urban function by integrating socio-economic information, address data and remote sensing. The geocoded address data forms the essential link that allows the integration of "real world data" as derived from remote sensing with non-spatial data such as company data from yellow pages. The use of Thiessen polygons proved to be valuable for linking the address points with parts of buildings. The final data base represents each building as a polygon in a GIS with its average height and its volume as physical attributes. In addition it is assigned a dominant function (residential, commercial, industrial, service, public) or a mixed use indicating the proportions of the different uses. Furthermore each building is linked to the companies that are located in it and its corresponding lines of business. For each building that has a full or partial residential use the estimated number of residents is given as well.