The real economy in Japan comprises two million firms and billions of links of supplier-customer and other relationships. We challenge to visualize such big data, specifically how firms are connected to each other. We use a dataset of a million of firms and millions of links of supplier-customer and ownership relationships in order to directly observe such a network structure. We discovered interesting features of the economic network. Firms are connected by the links into tightly-knit groups with high density in intra-groups and with lower connectivity in inter-groups, namely by community structures. We found that the features of commu-nity structures are specific to individual industrial sectors, such as manufacturers, retails and wholesales, infra-structure-related sec-tor, health-cares, and so forth. (Figure 1, the representative image attached to this paper is an example of graph drawing in which a million nodes are located with colors representing different indus-trial sectors). In the network, we also visualized how bankruptcies or failures of firms took place during one year right after the Great East Japan Earthquake in 2011.

Presenter(s)

Yuji Fujita , Turnstone Research

Yoshi Fujiwara , University of Hyogo

Wataru Souma , Nihon University