A Personalized Taxi Matching Algorithm for Ridesharing

This paper proposes a personalized taxi matching algorithm (PTMA) for ridesharing to meet the needs of different passengers. Our algorithm is divided into two phases: first, filtering to obtain the candidate set of taxi to be matched within the constraints of detour; second, using personalized taxi matching algorithm to insert the request into the candidate taxi, then calculating the score of each taxi and selecting the taxi with the highest score to match with the request in return. The experiment results show that the PTMA algorithm performs better in both the matching success ratio and the utility score compared with the T-share algorithm.