Discovering spatial patterns of fast-food restaurants in Jakarta, Indonesia

This study aims to discover the spatial patterns of fast-food restaurants in Jakarta, an Indonesian megacity. The location of fast-food restaurants tends to cluster with the Nearest Neighbor Index of 0.41. The average surface distribution (estimated by Kernel Density Estimation) shows at least 20 locations that have densities between 4-7 fast-food restaurants. The co-location rules, as the results of transaction-based co-location pattern mining, revealed six main activities that affect the location behavior of fast-food restaurants, namely leisure time/shopping, traveling, education, religious activities, health activities, work activities. These refer to 11 minor attributes of public facilities with the Participation Index more than 50% and at a distance less than 1 km. Foodservice providers can employ the co-location rules in determining locations for their brand expansion to other regions. The research framework to mine up information based on spatial data can be utilized both for business and academic needs for other purposes.