Lager availability of high-frequency and high-dimension firm-level micro data with many observation （i.e., “big data”） and developments of empirical techniques suitable for handling such data （e.g., machine learning） in recent years allow economist to conduct various empirical analyses also applicable to actual economic problems in policy design and business planning. In the contexts of prediction and causal inference, which are the two major empirical approaches to such micro data, it is further useful to combine spatial/geographical data such as location of enterprises, transaction of real estate, and development of infrastructure etc. Such joint employment of firm-level micro data and spatial/geographical data contributes to, for example, better identification of new entrants and exits from markets （esp., in the case of establishments）, expanding the dimension of firm attributes and frequency of data, and better identification of exogenous shocks to each firm.