GIS based automated valuation models – a genuine solution for real estate valuation in Romania

In the latest period, the usage of statistical models in the valuation of real estate properties became very popular among appraisal specialists, bankers and researchers from this field. Many Automatic Valuation Models (AVM) using different statistical and mathematical models like regression, neural networks, and fuzzy logic were proposed and developed with the purpose of assessing the value of a property. Also Computer-Assisted Mass Appraisal (CAMA) became popular for mass valuation of the real estate in fields like taxation or banking sector. The main weakness of an AVM and a CAMA is the absence of the locational analysis of the real estate; even if the spatial component is a main characteristic in establishing the commercial value of a certain property. Geographical Information Systems may be used in order to improve the accuracy of the Automated Valuation Model (AVM) or Computer-Assisted Mass Appraisal (CAMA) by combining the statistical evaluation methods with the spatial locations on property value. The cadastral maps as part of a geographical information system are essential to correct the mathematical models used in Computer Assisted Mass Appraisal. In this article, we analyze the applicability of geographical information system in property assessment, evaluating if the GIS automated models can be a solution for real estate valuation using a case study from the city of Oradea, Romania.