Strategic players for identifying optimal social network intervention subjects

Abstract

We present a method whereby social network ties are used to identify behavioral leaders who are situated in the network such that these individuals are: 1) able to influence other individuals who are in need of and most receptive to intervention, thereby optimizing the impact of the intervention; and 2) not embedded with ties to individuals that are likely to be behaviorally antagonistic to the intervention or that would compromise the optimal impact of intervention. In this study we developed a method that we call Strategic Players, which is a solution for identifying a set of players who are close to a target subset of the network (i.e., the target group), and far away from the subset we wish to avoid (i.e. the avoid group), where the proximity to either the target or avoid group may be facilitated by network members who are in neither group (i.e. the neutral group). This solution seeks to maximize the diffusion of the behavior to the target group while minimizing contact and influence to the avoid group. We apply this method to two different social networks, and one simulated social network.