LAUSR

It is critical for urban planners and real estate developers to understand how the built environment and house characteristics are valued in housing market. However, this problem is challenging because of the existence of the submarket effect resulted from the heterogeneity nature of city. In this paper, we propose a probabilistic approach to residential property hedonic valuation problem modeling the full scope of submarket effect based on built environment and house characteristics. Specifically, we introduce a latent variable representing housing submarket and model both of the submarket criteria and hedonic price model(HPM) into a Bayesian network. Utilizing the probabilistic dependencies in the Bayesian network, our model is able to capture the full scope of the submarket effect. Furthermore, to analyze the relationship among the discovered submarkets, we propose a probabilistic hierarchical clustering method to infer the hierarchical structure of housing market. In particular, we perform Bayesian hypothesis testings to find the most similar submarkets and agglomerate submarkets step-by-step, thus revealing the hierarchical structure of housing market. Finally, we conduct comprehensive experiments in the housing market of Nanjing which is a metropolis in eastern China. The experimental results demonstrate the effectiveness of our proposed modeling method.