A personalized event–participant arrangement framework based on user interests in social network

In recent years, online event-based social networks (EBSNs), such as Meetup and Whova, which offer platforms for users to plan, publish and arrange events, have gained popularity and rapid development. A fundamental task of managing EBSNs is to recommend a personalized and suitable sequence of events for potential users. However, none of the existing approaches consider both personalized recommendation and event arrangement at the same time. In addition, as to event-to-user recommendation, there are many existing methods, but they neglect individual differences between users and the differences between event sets that users can participate in. As to event arrangement, the different requirements of service providers are ignored, such as time efficiency, user experience, etc. In order to address these problems, in this paper, we first define the problem of Personalized Event–Participant Arrangement based on user interests (PEPA), which contains two parts: personalized recommendation and event–participant arrangement. For the personalized recommendation, we propose a personalized recommendation framework, which treats each user differently and computes the interest value toward events for each user, and we put forward a concept of user curiosity for the first time to overcome the problem of neglecting the differences between event sets for each user. For the event arrangement, we devise a Pruning-based greedy algorithm and two improvement strategies called Conflict-set strategy and Relaxation strategy to improve the runtime efficiency and the user experience respectively. Finally, we conduct extensive experiments on real and synthetic datasets to verify the efficiency and the user satisfaction of the proposed algorithms.