DBPapers
DOI: 10.5593/SGEM2014/B22/S9.017

CLUSTER ANALYSIS OF LAND PROPERTIES FOR THE PURPOSE OF MASS APPRAISAL

M. Maleta; E. Bielecka
Wednesday 1 October 2014 by Libadmin2014

References: 14th International Multidisciplinary Scientific GeoConference SGEM 2014, www.sgem.org, SGEM2014 Conference Proceedings, ISBN 978-619-7105-11-7 / ISSN 1314-2704, June 19-25, 2014, Book 2, Vol. 2, 129-136 pp

ABSTRACT
The aim of this research was application of data clustering, especially k-means clustering algorithm, used for grouping land property for the purpose of elaborating tax map. Morphometric characteristics such as parcel’s size, parcel’s shape, area, and type of land use have been used as exogenous variables whereas transaction price per 1 square meter was adopted as endogenous variable. Since the classification result (kmeans method of classification) depends on assumed a priori number of class therefore to estimate number of clusters both, agglomeration method and Ward’s algorithm have been used. Land properties grouping has been done for Stoczek Lukowski municipality, located in south-west part of Poland. Data for analysis were taken from lands and buildings register and from registry of real estate prices and values. Due to the fact that real estate prices were prices of the years 2002-2012 in the preliminary stage of the study they were updated using multi-linear regression method. Land average transaction price for real estates of each group is similar, this made mapping of the property value possible. Such map is a decent source of information for both, investors and local government responsible for the management of a municipality area in a sustainable way.

Keywords: grouping properties, land value map, mass appraisal