The use of data by smart systems for price forecasting in the context of building customer relationships on the Lublin real estate market

Abstract

This article aims to show how to design a functional application as an intelligent tool for forecasting real estate prices, among others, based on historical data. Real estate dealers or clients in this market are rarely aware of the possibility of using such a tool in the real estate valuation. This paper presents how the artificial neural network can be used for this purpose. The selected area of the Lublin real estate market has been covered by research carried out for this study. As input data for the smart system, historical data of transactions and variables characterising real estate were used, such as location of the property, its condition or size. Using the developed model, in the neural network, the sensitivity of the real estate price to selected features was carried out. It was also indicated how the use of this tool can facilitate purchasing decisions and customer relationship management on such a market.