Patterns of success and cooperation in complex networks

Abstract

I look around about these cooperation and success. First, through the prisoner’s dilemma, it is started from the most microscopic part of the cooperation. It is confirmed that cooperative strategies among 25 strategies with single step memory can survive in replicator dynamics simulation with mean-field approximation and construct win-lose directed network between strategies. Especially, among these 25 strategies, there are win-lose circulations like rock-scissors-paper in the network. And it is investigated a dynamic motif which represents all possible win/lose relations including circulations among three strategies. In particular, It is investigated the payoff dependent fine phase structure of networks using relation changing between strategies. Finally, It is simulated an emergence of cooperation with Pavlovian-walk in a numerical model which mobile agents with a reputation-based strategy. It is an opportunity to understand the ‘emergence’ of cooperation through the prisoner’s dilemma, and it is looked at how GitHub achieves success and cooperation in real data. First, it is defined an active user and an active repo, and it is measured some popularities of a project as a measure of success of open source. It is defined an openness how much repository is opened to contribute of other users and examine its implications. It is measured burstiness and memory of repository using inter-event times. It is constructed a network based on collaboration on GitHub, and measure various centrality on networks. It is examined the implication in correlation between all metrics for GitHub system. Finally, it is suggested a growing network model which describing collaboration in GitHub.