Weight Impact on Comparative Evaluation of Topographic Data

Abstract

The paper addresses the problem of weighting in an analysis that supports the selection of a categorical data set according to user needs. Using the Relative Change (RC) of the Compound Correspondence Index (CCI), it is shown that weights have a significant impact on user choice – reaching extreme values in both urbanized and forested areas. Decreasing the weights from 0.25 to 0.17 in forested and built-up areas resulted in the maximum variations that were seen in the hot spot maps, with cold areas generally corresponding to built-up regions and hot areas to forested areas. The analysis covers seven counties that are located in different regions of Poland: Pomerania, Podlasie, Mazovia, Greater Poland and the Beskidy Mountains.