Spatial Modeling and Geovisualization of Rental Prices for Real Estate portals

State Of The Art Of Analyzing And Mapping Rental Prices

To determine how real estate portals analyse and map rental prices, 32 real estate portals were examined by 14 criteria by applying statistical and visualization methods. In addition to this examination, and according to a designed standardized questionnaire, five 45 minutes interviews with renowned German real estate agents and real estate portal experts were conducted. The analysis examined the 32 largest real estate portals by unique web access (Alexa, 2013). Real estate portals have replaced newspapers as the go-to medium to search for rental property. Just to demonstrate the dimensions, the Germany portal Immobilienscout24 constantly stores 1.5 million rental properties in its database (Immobilienscout 24, 2012). The fact that all these offers are geocoded makes them a great, but so far insufficient used source for spatial real estate analysis and visualization.

Summarizing the results of the portal examination, it can be stated that the most real estate portals offer real estate market indices for clients as banks, investors and public authorities. Those indices neglect the location of real estates and are built by means of the non-spatial statistical method of simple descriptive statistics (mean, median) or by applying hedonic regressions. A common procedure is to map descriptive statistics- or hedonic regression results onto non-appropriate reference geometries, e.g. on ZIP code areas or on city district levels. Trulia.com for example maps the median rental price by ZIP code (cf. Figure 1). The resulting rental price maps show a distorted distribution of rental prices in space, which does not correspond to the real distribution of price in space.

Figure 1.

Example of the US real estate’s portal Trulia of mapping a median rental prices on the geometries of New York’s’ city ZIP code areas (Trulia, 2016)