Controlling the preferential attachment evolution in the limit of a large number of agentsдоклад на конференции