Title: Real Estate Valuation Using Data Mining Software
Abstract: This paper presents an innovative method for real estate valuation. The most common method of real estate valuation relies on comparisons with similar structures, cost and yield analysis. The innovative method proposed in this paper, called "historical market price", uses mathematics, statistical and database-founded algorithms for valuation. Input data comes from specialized software which systematically gathers, analyses and valuates data connected with real estate market development. Every six months the database expands by more than 650,000 price offers (advertisements) on purchase or rent of flats, houses, commercial buildings and allotments. This method can be used for valuation of real estate which was purchased previously if rough data is known about the original transaction (especially purchase price). The method suggested in this paper is grounded in true and exact information taken from the last purchase. It also identifies structural and technical differences in the condition of given real estate, as well as changes in price levels for the given location (street, neighborhood, city, region, etc.). Current valuation methods are not capable of taking into account these historical, yet exact and valuable pieces of information concerning real estate. Current valuation methods are based solely on current data and present state of the real estate market. This approach may be vulnerable to manipulation by either party of the purchase contract.