An interaction-aware approach for social influence maximization

Abstract

Microblogging networks are considered a great source of social influence. One of its characteristics is their high dynamism. This fact produces that influential users continuously change according with time and topic. Several social networks metrics have been defined to rank influential users. However, these metrics fail to capture the dynamism of microblogging networks. For this reason, we propose an approach based on Credit Distribution model to identify the influential users of a microblogging social network by performing an online analysis of the users’ interactions. Moreover, we present a comparison of our approach with well-known metrics used for influencers ranking. The experiments were carried out in Twitter during sport events (football matches) and new product (video games) launchings. The results showed that our approach outperforms the metric-based rankings in terms of the influence spread. This confirms the importance of being updated for identifying influential users.