Information Selection and Delivery Algorithm for Delivering Advertisements Suitable for the Pedestrians Present at a Particular Site

We consider a service that shows an advertisement suitable for the preferences of the pedestrians present at the site where the display is installed. In order to evaluate the effectiveness of this service by a simulation that reflects the reality well, we propose evaluation functions that take account of the line of sight of moving pedestrians. The evaluation functions incorporate two independent variables: Category and Producer. Based on this evaluation functions, we also propose an algorithm of multicasting an advertisement to different sites in order to deliver advertisements efficiently over a network while ensuring a high advertisement relevance rate. The evaluation results show that the proposed algorithm always scores a higher advertisement relevance rate than conventional advertising media. It is also shown that the proposed algorithm can reduce the delivery time by about 37%, and the volume of data delivered from the server by about 39% compared with unicast delivery.