Author(s)

J M Choi & Y Asami

Abstract

Marketing analysis is an indispensable step to establish strategies for the condominium business. In this context, it is crucial to grasp the potential customers’ behavioral characteristics as \“rule” or \“pattern” extracted from historical data. In this paper, the earlier straightforward approach is outlined and the extension based on conditional contingency table is demonstrated as an alternative to extract unexpected rule patterns from a database with relatively few transactions but a huge number of attributes. The paper shows that this approach can better summarize information targeted to unexpected patterns than the earlier method using data mining methodology. For the validity of the proposed method, about 800 condominium purchasers in Tokyo metropolitan area are being analyzed and compared with the result of the earlier method. 1. Introduction Most of major real estate companies in Japan have been collecting customer data and building massive data warehouses. Few companies however, have been able to realize the actual value stored in it. The question these companies are asking is how to extract this value. One of the possible answers for this could be data mining. Data mining technology has increasingly been fascinating marketers as it can provide valuable hidden business intelligence from historical data. The method can be soundly employed in the context of a need to better understand customers, market segmentation, and to quickly respond to their individual needs and wants. Nevertheless, the actual implications of rule mining in the housing market are very scarce. A couple of reasons for this can be cited, one being that it is often very hard for a researcher to obtain detailed information about the private housing sector, and even after successful gathering of this information, but owing to business confidentiality there are usually limitations on opening them to the

Keywords