Explaining prices by linking data: A pilot study on spatial regression analysis of apartment rents

Abstract

This paper reports a pilot study where we link different types of data for explaining prices. In this study, we link the apartment rent data with the publicly accessible location data of landmarks like supermarkets, hospitals, etc. We apply the regression analysis to find the most important factor determining the apartment rents. We claim that the results of this type of spatial data mining can enhance the user experience in the apartment search system, because we can indicate a rationale behind pricing as additional information to users and thus can make them more confident in their choices.