Speed dating: An algorithmic case study involving matching and scheduling

Abstract

In this paper we formalize and solve the speed dating problem. This problem arises in the context of speed dating events, where several people come together to date each other for a short time. For larger events of this type it is not possible to have each possible pair of persons meet. Instead, based on forms filled out by the participants, the organizer of such an event decides in advance, which pairs of people should meet and also schedules the times of their dates. Moreover, since people pay for participating in such events, aside from the overall quality of the dates, it is important to find a fair schedule, where people from the same group (e.g., all women) have a comparable number of dates. We model the organizer's problem for speed dating, study its complexity and design efficient algorithms for solving it. Finally, we present an experimental evaluation and show that our algorithms are indeed able to solve realistic problem instances within a reasonable time frame. © 2011 Springer-Verlag.