In search of influential event organizers in online social networks

Abstract

Recently, with the emergence of event-based online social services(e.g. Meetup), there have been increasing online activities to create, distribute, and organize social events. In this paper, we take the first systematic step to discover influential event organizers from online social networks who are essential to the overall success of social events. Informally, such event organizers comprise a small group of people who not only have the relevant skills or expertise that are required for an event (e.g. conference) but they are also able to influence largest number of people to actively contribute to it. We formulate it as the problem of mining influential cover set (ICS) where we wish to find k users in a social network G that together have the required skills or expertise (modeled as attributes of nodes in G) to organize an event such that they can influence the greatest number of individuals to participate in the event. The problem is, however, NP-hard. Hence, we propose three algorithms to find approximate solutions to the problem. The first two algorithms are greedy; they run faster, but have no guarantees. The third algorithm is 2-approximate and guarantees to find a feasible solution if any. Our empirical study over several real-world networks demonstrates the superiority of our proposed solutions.