Collaborative activity-based ridesharing

Abstract

A new ridesharing model called collaborative activity-based ridesharing is proposed to enhance not only overall matching rates but also the matches between preferred ride partners. By coalescing the merits of two recently suggested innovative ridesharing models – the social network-based ridesharing and the activity-based ridesharing – the new model leverages people's preference to their social networks and the space-time flexibility of daily activities to improve the matching outcome. The capabilities and advantages of the proposed model are justified by a group of agent-based simulations in a realistic study area. The influence of geography on the match outcome is discussed in particular.