Entropy model analysis on the commodity and information flow structure

This study focuses on the structural difference between commodity and information flows, being based upon the OD data of numbers of commodities and telephone calls among 56 areas in the Kinki district. Firstly, the paper investigates the distribution of total numbers of commodity and information flows going to and from each area by applying the idea of order statistics. Then the connection between areas, which is defined by the amounts of flows from one area to another, can be analyzed through graph theoretical approach. Gravity model analysis can be applied to express the flow from one area to another in terms of the total flow to and from the area and the distance between these two areas. Thus the elasticities of the amount of flows with respect to the total flows and the distance can be obtained. Finally, as a main part of the study, entropy model approaches are proposed in order to explain the structure of the commodity and information flows. Thus the total flow going from each area can be formulated as a one factorial path type entropy model, while the flow from one area to another as a generation constrained entropy model. Furthermore, the correspondence between commodity-information flow and entropy maximization process is shown theoretically and experimentally. [In Japanese.]