Abstract:
|
Foot traffic is one of the most important factors in location planning and location-based marketing. In this paper, we analyze the Foursquare foot traffic data in a network framework. Firstly, we build a directed location-location network to connect those locations that are commonly visited in one trip. We then developed a linkage prediction model to predict how likely two locations will be linked together. The linkage prediction model takes into consideration of the distance, industry, number of visitors, number of the competitors and etc. Moreover, we developed a score to measures the closeness of the two locations in the foot traffic network. We show the benefit of using it in real applications in location planning and location-based marketing.
|