Scientific Collaboration Analytics based on Multi-Player Game Theory

Exploring behavioral patterns amongst scholars in Academic Social Networks (ASNs) has attracted significant attention in the field of social computing in recent years. Scholars are faced with a dilemma in which they either opt to cooperate with others or to spend their limited resources for their own individual gain. These interactions and decisions can be modeled using Game Theory. Recent research has shown that favorable longterm strategies are characterized as being reciprocal, forgiving, and flexible. Little focus has been given to the phenomenon of Upstream Reciprocity (UR), which occurs when a person who has just received help, in turn, offers help to another, in a `pay it forward' manner. This paper provides a framework to model scholarly interactions of ASNs using an iterative multi-player game and investigates the occurrence of UR amongst players. The presence of UR amongst scholars is observed and shows to possess a higher publication and citation count by those who adopt it. Moreover, cascades of UR have been identified to replicate throughout the network, thereby increasing further adoption of this behavior. More importantly theoretical proof is provided that suggests that UR has the potential to become an Evolutionary Stable Strategy (ESS).