研究期間：10108~10207;The purpose of this proposal is to automate pricing model for real estate using computational intelligence algorithms. A database will be established at first, integrating a GIS system and actual trade data. The database is expected to include estimate datasets of over 5,000. As a result, we will identify and quantify attributes that affect buyers’ pattern by the use of the analytic hierarchy process (AHP) technique so as to bring up customers’ viewpoints for practitioners and to set the inputs for the next stage. Using such inputs, the prediction model founded on the support vector machine (SVM) concept will promisingly yield outcomes forecasting real estate prices. Third stage will adopt a novel clustering algorithm named swarm-inspired projection (SIP) to explore data characteristics and clustering features. The results obtained from SIP will be compared to those from SVM. The anticipated contributions lie in the identification of attributes by customers’ viewpoints, prediction and pricing of real estate, and the comparison between classification and clustering in solving the problem of real estate pricing.