TW201613396A - Crowd potential information generation method - Google Patents

Abstract

This invention discloses a crowd potential information generation method which analyzes the connection positioning data of a mobile device at a base station and makes use of a hierarchy geographical clustering method, periodic movement people amount prediction module, temporal correlation algorithm and external information introduction to predict the crowd amount and movement pathway. The hierarchy geographical clustering method is used to cluster the base stations, in which base stations close to each other are assigned to the same cluster, so as to solve the problems where mobile devices, due to high distribution density of base stations in metropolitan areas, switch among several identical base stations within a short period of time and solve the problem of excessively large geographic area of clusters due to low distribution density of base stations in suburb, such that, after clustering, the mobile devices can be positioned in the constant cluster area scope to achieve local crowd potential analysis. The statistical operation result of the local crowd is used in conjunction with time flow calculation function to establish the periodic movement people amount prediction module to predict the future crowd amount. The temporal correlation algorithm can calculate the crowd movement path and possibility and introduce external information to find out the external factors that influences the crowd potential information, thereby increasing the accuracy of crowd potential analysis.