On trajectory prediction in indoor retail environments for mobile advertising using selected self-histories

In this paper, we propose a new trajectory prediction algorithm, 3 Indoor Trajectory Analysis (3-ITA), to analyze the different trajectories of a shopping centre visitor, and to predict the trajectory with high probability that the visitor could use on a specific date. Here, a trajectory is a sequence of stores visited by a user, not necessarily the actual physical path/walk taken by the user when visiting the stores. 3-ITA is capable of predicting how many stores will be visited by that visitor, what those stores are, and the order of visiting each one of the predicted stores. Moreover, 3-ITA is working on an individual level. Every trajectory-pattern analysis and every predicted trajectory is exclusively related to the profile of a registered visitor. In addition, we developed a prototype system called Ad4Me. Ad4Me is a personal trajectory-pattern-aware pervasive system for mobile advertising in retail environments. Ad4Me will use the predicted stores from 3-ITA to detect the interests of a shopping centre visitor, in order to generate an adaptive list of relevant ads to that visitor. The experimental results show the high prediction accuracy of our algorithm.