D-Taxi: Adaptive area recommendation system for taxis by using DiRAC

In this paper, we present a recommendation system, D-Taxi, which informs taxi drivers on where to find the next passenger using the latest picking-up/dropping-off and positional information from other taxis. D-Taxi works in terms of the global optimization of the total vacancy time of taxis without any deep analysis of the historical logs by using our control mechanism named DiRAC. D-Taxi can adapt to the real-time situation such as an unexpected increase of passengers. The experimental results show that the proposed system can adapt to the aforementioned situation, and can improve the occupancy rate by 20% compared with a greedy algorithm.