Aneta Chmielewska, Małgorzata Renigier-Biłozor (p)

Marek Walacik

Residential properties are a major component of the environment and economy and a key element for the quality of human life. Faced with disruptive ideological and technological changes in the world, real estate analysis has also become a key research problem for many academic centers and private institutions. Due to the complex nature of properties, they are one of the most difficult and troublesome subjects of analysis. Given the rapid advancements in competitive automated analytical models, the problem of data representative sample selection may prove to be a very wide-reaching subject. The aim of this paper was the assessment of the particular soft computing methods’ (e.g., Self-Organizing Maps, Rough Set Theory) usefulness for selecting a representative property model. The obtained results confirm that the use of these methods leads to the creation of a model that enables a more reality-based view of the uncertainty and imprecise residential environment.

Simon Thaler (p), David Koch

Malgorzata Renigier-bilozor

Assessing the condition of a property is an important part of determining its value, but also the location which is partly determined by the immediate surroundings of a house is of high importance. To assess the condition of a house, an appraiser evaluates certain (exterior) characteristics according to a scheme in which each characteristic is weighted for the final condition classification. Such a condition assessment can also be made from pictures of the exterior which naturally show features of the surroundings but do not require an on-site visit. Therefore, this article examines the extent to which 102 students, as a sample of the general population, evaluate the condition of 100 single-family houses pictures of the exterior. Four of the pictures have been manipulated, the rest is controlled for different features that are visible in the surroundings. The main result is that the wisdom of a crowd assessment of the condition significantly increases the adjusted r-squared in a hedonic model compared to all single condition assessments. Furthermore, the focus of the scrutiny is on the effect of the surroundings on the assigned class. Despite of the fact that 66% of the mode answers of all the students correspond to the assessment of experts we also find that e.g. the existence of a pool biases the assessment of the participants to a too positive class assignment. Furthermore, it seems that the amount of greenery as well as the weather conditions in the background bias the assessment of students. Colour or black and white seem not to influence the assessment.

Geoffrey Turnbull, Arno Van Der Vlist (p)

Simon Thaler

We investigate effects of crimes in housing markets. Using unique data on gated and non-gated subdivisions, we identify capitalization effects of crimes on housing prices. For this, we decompose data on crimes in gated and non-gated crimes for gated and non-gated property. Data from Orange County Florida over 2010-2013 indicate that crimes are associated with lower house prices. We find a mean effect of 2 per cent lower house prices for every crime within 520 feet and 180 days prior to the house transaction. For non-gated property we find a mean effect of 1 per cent lower house prices for nearby non-gated crimes. For gated property we find a mean effect of 8 per cent lower house prices for every nearby gated crime.

Marek Walacik (p), Artur Janowski

Arno Van Der Vlist

Every real estate market analysis cannot be conducted without a necessary identification of sub-markets, or market unit areas. The identification is vital in relation to price formation phenomena and factors affecting real estate value that from the elementary (principal) assumptions are (must be) homogeneous. The problem is up-to-date and vital in wider international discussions as evidenced by numerous publications. Definition of the homogeneity of the area related to the property markets is strictly connected with the aim of the property analyses. In property market analyses and valuation, the most popular issue when the submarkets and homogeneity of the area should be defined is comparable property transactions selection. From this point of view, a homogenous unit is an area in which a certain number of real properties subject to an appraisal demonstrate the same impact of the location on their value. In other words, all the real properties located in a given elementary area formally do not differ in terms of their locality. Considering the research problem presented above, the main objective of the presentation is to provide the current state of art solutions overview and original solutions formulated by the author.