1. A novel game theoretic approach for modeling competitive information diffusion in social networks with heterogeneous nodes
- Author
-
Farnaz Barzinpour, Alireza Aliahmadi, Mehrdad Agha Mohammad Ali Kermani, and Seyed Farshad Fatemi Ardestani
- Subjects
TheoryofComputation_MISCELLANEOUS ,Statistics and Probability ,Computer Science::Computer Science and Game Theory ,Mathematical optimization ,Computer science ,02 engineering and technology ,01 natural sciences ,010305 fluids & plasmas ,symbols.namesake ,0103 physical sciences ,0202 electrical engineering, electronic engineering, information engineering ,Social network ,business.industry ,Node (networking) ,Stochastic game ,Statistical and Nonlinear Physics ,Maximization ,Incentive ,Nash equilibrium ,Best response ,symbols ,Topological graph theory ,020201 artificial intelligence & image processing ,Epsilon-equilibrium ,business ,Mathematical economics ,Game theory - Abstract
Influence maximization deals with identification of the most influential nodes in a social network given an influence model. In this paper, a game theoretic framework is developed that models a competitive influence maximization problem. A novel competitive influence model is additionally proposed that incorporates user heterogeneity, message content, and network structure. The proposed game-theoretic model is solved using Nash Equilibrium in a real-world dataset. It is shown that none of the well-known strategies are stable and at least one player has the incentive to deviate from the proposed strategy. Moreover, violation of Nash equilibrium strategy by each player leads to their reduced payoff. Contrary to previous works, our results demonstrate that graph topology, as well as the nodes’ sociability and initial tendency measures have an effect on the determination of the influential node in the network.
- Published
- 2017