Distance-aware event recommendation in event-based social networks

Event-Based Social Networks (for instance, Meetup.com) allow users to create, promote, and share with other users upcoming events of any kind. Event-Based Social NetWorks provide recommendations to users to assist them in finding those events that best match their preferences. However, the event recommendation problem raises several issues that are different from other domains, such as books or music. Events rapidly disappear, users' preferences quickly change over time, and direct feedback does not exist, since events have not taken place yet. In this paper, we propose two context-aware event recommendation algorithms, which consider information about distance between the user and the events. Additionally, we compare the effectiveness of different algorithms in the event recommendation task, considering the feedback coming from RSVPs (Répondez S'il-Vous-Plaît, meaning please respond). We validate our proposal on two datasets extracted from Meetup, considering standard accuracy metrics. Results show that hybrid version containing collaborative and distance-aware algorithms ranks the best among the tested algorithms.