In recent years, many retail establishments in Japan have attached amusement functions along with selling goods such as movie theaters, game centers, and themed food courts. A reason for these trends is a drastic diversification of consumer's attitudes and preferences. In order to acquire new and more loyal customers, shops must grasp diversified consumer demands and provide for their customers the unique characteristics other stores cannot offer such like amusement facilities. Similarly, the functions of open cafes, events, street performers, and music performances in an open urban space also have been reevaluated because of their important roles for attractiveness of city. Exploring what kinds of consumers are attracted by what kinds of amusement functions is worth investigating further in urban studies. For the purpose, we address the problem to extract critical attributes of frequent visitors for different kinds of urban commercial complexes for investigating the roles of amusement functions to attract customers. First, we apply C4.5 algorithm to visit frequency data for a themed noodle food court set up in a commercial complex for exploring what kinds of customers are attracted by the amusement function of the themed noodle court. The C4.5 is the representative algorithm of decision trees, which are classification algorithms that serve to discover significant relationships between explanatory attributes and a given predicted class in the dataset. We adopt customer's profiles as explanatory attributes and frequent shopper as a predicted class. Thus, using C4.5, we extract the rules to explain critical profiles of frequent shopper for this noodle court. Furthermore, we apply 1R algorithm, which produces one-level decision tree, to visit frequency data for different kinds of 8 urban commercial complexes for exploring what kinds of profiles characterize frequent shoppers for these complexes.