En uygun konut seçimi problemine mekânsal karar destek süreci ile alternatif bir yaklaşım

Abstract

Son yıllarda artan internet kullanımı her alanda olduğu gibi taşınmaz sektörünün de vazgeçilmez bir parçası haline gelmiştir. Günümüzde taşınmaz malikleriyle –taşınmaz edinmek isteyenlerin iletişiminde kullanılan en yaygın ve etkili araç taşınmaz web uygulamalarıdır. Talebin ve gereksinimin üzerinde bulunan konut sayısı, insanların yoğun yaşamlarında ilgilendikleri konutların hepsini fiziksel olarak görecek zamanlarının olmaması vb. nedenler en uygun taşınmazın belirlenmesini ve güvenilir taşınmaz seçiminin yapılmasını zorlaştırmaktadır. Bu durum nitelikli ve güncel veriye ulaşma gereksinimi ile birlikte taşınmaz/emlak web sitelerini de daha hızlı, farklı ve kaliteli çözümler geliştirmeye yönlendirmektedir. Bu çalışmada konut arayan kişiler öncelikli olarak, kullanılan web sitesi üzerinden konum, oda sayısı, metrekare, fiyat ve tarih aralığı gibi aradığı taşınmazın fiziksel özelliklerini sisteme girerek, taşınmazların fiziki yapılarıyla örtüşen ilanlar filtrelenmiş ve bu ilanların harita üzerinde gösterimi sağlanmıştır. Sonraki aşamada kullanıcının önem verdiği veya kendi ihtiyaçlarına uygun çevresel ölçütlerin seçimi ve değerlendirilmesi ele alınmıştır. Çalışma kapsamında çevresel ölçüt; diğer bir deyişle erişebilirlik ölçütleri olarak sağlık tesislerine yakınlık, yeşil alan ve parklara yakınlık, ulaşım olanaklarına yakınlık, eğitim tesislerine yakınlık ve dini tesislere yakınlık olarak kabul edilmektedir. Taşınmaz sektöründe çalışanlarla yapılan anketlerde konut seçimine etki eden çevresel ölçütler ve ağırlıkları belirlenmeye çalışılmıştır. En uygun konutun seçiminde kararı etkileyecek ölçütlerin sayısal olarak anlamlandırılması için Çok Ölçütlü Karar Destek yöntemi kullanılmıştır. Çok ölçütlü karar destek sürecinde özellikle mekânsal erişebilirliğe ilişkin ölçütler öne çıkarılmıştır. Bu çalışma kapsamında konut arayanların web sitesi üzerinden seçtiği çevresel ölçütler ya da diğer bir ifadeyle kullanıcının karar vermesi için önemli bulduğu ölçütler yakınlıklarına göre matematiksel işlemlere sokularak ilgili taşınmazlar için bir puanlama yapılmakta, yapılan puanlama sonucunda konut arayanlar için öncelikle ölçütlere göre konutlar sıralamaya sokularak karar destek süreci gerçekleştirilmektedir. Kullanıcının kendi önceliklerine göre seçmiş olduğu erişebilirlik ölçütlerinin, fiziksel ölçütlere göre filtrelenmiş konut sonuçlarının her birine olan mesafesi ile ağırlıklandırılarak sonuç kümesi elde edilmiştir. Ayrıca kullanıcının belirlediği her bir erişebilirlik ölçütünün ve birbirlerine olan ağırlıkları da kendi aralarında ikili karşılaştırmaya sokulmuştur. Elde edilen sonuç kümelerinin birleştirilmesi ve anlamlandırılmasıyla her bir konut için puanlandırma yapılarak kullanıcıya karar destek sağlanmıştır. Konut çıktılarının hem iki boyutlu hem de üç boyutlu haritalar üzerinden görselleştirilerek karar destek süreç sonuçlarının daha iyi anlaşılması sağlanmaya çalışılmıştır.

In recent years, the increasing use of the internet has become an indispensable part of the real estate sector as it is in every field. Today, the most common and effective tool used in the communication of real estate owners and customers, which is one of the components of the real estate sector, is real estate web applications. The increasing number of houses above the demand makes it difficult to choose reliable real estate. In this case, the need to reach qualified and up-to-date data, as well as real estate website aims to develop faster and quality solutions. In this study, first of all, people which are looking for real estate; by using the website, the properties of the immovable properties such as location, number of rooms, square meters, price and date range will be entered into the system and the advertisements overlapping with the physical structures of the real estates, will be visualized on the map. As the next step, user's regarding environmental criteria had dealt that selection and evaluation of the that the user attaches importance. Environmental factors of a real estate can be expressed as earthquake risk zone, flood and flood risk, noise pollution, green area amount and proximity, proximity to transportation, etc. In the selection of real estate, important environmental criteria were determined with the survey conducted with the employees in the real estate sector and Multi-Criteria Decision Support methods were used to quantify the factors affecting the decision. On the web site, environmental factors that were chosen by the property seekers or in other words, factors that user finds important for the decision had processed and the numerical equivalents or scoring of the related real estates are given to the user as outputs for each house. As a result, decision support was provided for the most appropriate real estate selection according to the physical and environmental factors that were determined by the user. Spatial decision making, like real estate choosing and spatial accessibility, is an extremely complicated operation of selecting the optimal alternative within a number of possible alternatives to reach a set of objectives based on a decision maker's subjective and contradictory options under restrictions. Solutions for these spatial decision problems involve extremely complicated spatial data analysis operations and constantly require advanced means to address physical conformity circumstances while considering the multiple socioeconomic and peripheral factors. The demand for advanced means providing sophisticated spatial modeling, assessment and visualization skills has enhanced over the last two decades in a number of implementation domains in order to adequately investigate and solve unstructured complicated employment issues with spatial perspectives. Spatial decision support systems (SDSSs) assistance spatial problem-solving and strategic decision-making actions by combine analytical/decision models with spatial functions. SDSS can be adequately used to manage a multicriteria analysis (MCA) and assemble it with a spatial location. The data are processed using geospatial information system (GIS) and multicriteria decision analysis (MCDA) methods to acquire information for making decisions. The main goal of the study is to develop a flexible Web-based MCDA framework, called EMEKLI, which is able to visually support the decision makers in performing Web-based decision-making processes in spatially related problems. In this study, the selection of the most suitable apartment (or a flat) has been considered as a spatial decision problem. EMEKLI can be accessed on the Web address of http://hakanemekl i.xyz/. To efficiently analyze and solve the problem of apartment selection, the study covers two main criteria: spatial aspect and house attributes. Non-spatial aspect includes the inputs such as unit price, house size, house age, storey, and number of rooms. Spatial aspect is related to what kind of services exists in the surrounding area of a property. To achieve that, measures of accessibility are used as criteria to supply MCDA. EMEKLI can provide the overall evaluation including housing itself and surrounding environment for candidate apartments according to the decision makers' preferences. The overall score for each candidate apartment is calculated, and then the apartments are listed according to their scores in descending order on the Web-based interface of EMEKLI. Three-dimensional (3D) geovisualization of the buildings where the candidate apartments are within assists the decision makers in understanding the reality of surroundings in order to help them judge the alternatives. Since different multicriteria decision-making (MCDM) techniques are applicable in different decision-making situations, a specific MCDM method is needed to be determined in order to meet the requirements of a particular problem, such as the selection of an apartment. In this study, the analytic hierarchy process (AHP) method has been employed into the problem of apartment selection. In this study, EMEKLI has been developed as a Web-based spatial decision framework to facilitate the decision-making process for apartment selection in the presence of different preferences and uncertainties among the decision criteria by integrating multicriteria analysis and spatial analyses to 3D visualization. The most of the real estate agencies' Web sites provide basic SQL query tools. They are not the Web-based spatial information system applications although they use map and location information. However, EMEKLI can be utilized for consultancy purposes in determining the most suitable apartment that locates the available apartments by weighting functions according to the decision maker's spatial preferences with computational and visual technologies. Since EMEKLI can identify the better candidate apartments than traditional sorting-based approach by considering the overall property attributes and surrounding facilities, this makes the proposed approach superior than the real estate agencies' Web sites. A case study of EMEKLI has been developed in the district of Sisli in Istanbul, Turkey. The framework developed in this study can be applied in other districts of the city and to the entire city or other cities worldwide. To achieve this in an easy manner, the different data sources such as network datasets and PoI databases would be needed. During the case study, it has been observed that transportation and PoI databases in the level of national, urban, and local are needed in a Web-based spatial service structure that could support high-level network analysis on the Web. Moreover, GIS-based real estate valuation services, real estate value maps, real estate appraisal value models can be integrated with analytical processing services of the real estate agencies' applications to determine real estates' market value and the future trend in property price. Such an approach can be used by the decision makers for consultancy purposes in order to distinguish real value and market price of an apartment. This is very important for the cities where the real estates have exaggerated prices particularly in mega cities like Istanbul. In the scope of this study, the distance measurement functions are simplistic without considering multiple travel modes, such as walking, bicycling, public transport, private cars. Additionally, further research can consider different methods of MCDA in the calculation of the spatial accessibility. Furthermore, more surrounding facilities like workplace can be considered in the scale of entire city using multiple travel modes by taking into account traffic congestion and business hours, the speed limits of network segments and the effect of turn restrictions and directionality (e.g., one-way street) in order to more realistically model real-world networks. In the future, the other subjective and objective factors can enhance the work, such as income, composition of population, security, price trend, transportation cost, delays in traffic congestion.