Implementing GIS in real estate price prediction and mass valuation: the case study of Nicosia District
Date Issued
March 2017
Author(s)
DOI
https://doi.org/10.1117/12.2280255
Abstract
When the European Commission, International Monetary Fund and European Central Bank arrived in Cyprus to assist for a sustainable solution on the crisis on the banking sector, one of the first things they ordered was a New General Valuation (a mass appraisal that would revalue all properties in Cyprus as on 1st of January 2013), that it would be used for taxation purposes. The above indicates the importance of property mass appraising tools. This task was successfully conducted by the Department of Lands and Surveys. Authors aim to move a step further and implement the use of GIS and GWR techniques to improve the results of the New General Valuation. On a sample of comparative evidences for flats in Nicosia District, GIS was used to measure the impact of spatial attributes on real estate prices and to construct a prediction model in terms of spatially estimating apartment values. In addition to the structural property characteristics, some spatial attributes (landmarks) were also analysed to assess their contribution on the prices of the apartments, including the Central Business District (CBD), schools and universities, as well as the major city roads and the restricted zone that divides the country into two parts; the occupied by Turkish area and the Greek area. The values of the spatial attributes, or locational characteristics, were determined by employing GIS, considering an established model of multicriteria analysis. The price prediction model was analysed using the OLS method and calibrated based on the GWR method. The results of the statistic process indicate an accuracy of 81.34%, showing better performance than the mass valuation system applied by the Department of Land and Surveys in Cyprus with accuracy of 66.76%. This approach suggests that GIS systems are fundamentally important in mass valuation procedures in order to identify the spatial pattern of the attributes, provided that the database is comprised by a sufficient number of comparable information and it is continuously updated.