Tag
This paper presents a contextual multi-armed bandit framework that learns individual spillover probabilities in social networks to optimize stimulated word-of-mouth marketing, achieving higher rewards by targeting connected users.