Loading...

Improving The Efficiency of Social Networks :a Game Theoric Approach

Maazallahi, Abbas | 2014

519 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 45637 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Habibi, Jafar
  7. Abstract:
  8. Modeling is a tool to indicate mathematical structure of data and their relations.Game theory modeling defines utility functions by rationality property of agents. In game theory each agent attempt to maximize his(her) utility function.Game theory modeling is generally define under multi-agent systems in which each player is an autonomous agent in the multi-agent system. If we consider the whole system as a single agent that has one particular goal, accomplishment of this goal could be mathematically modeled. Objective function is the measure index to show how a system achieves its goal.Utility function in game theory and objective function can be conceptually correlated with each other in three moods. In first one, these two functions are positively correlated so that growth in utility function will lead to increase in objective function. In next one, there is no meaningful and definable relation between these two functions and in third one, the functions are negatively correlated which means that increase in utility function will decrease objective function.
    By specification, social networks are multi-agent systems which can be used to achieve some goals such as information propagation and communication reinforcement. Social networks (as a multi-agent system) have a target which can be modeled mathematically named objective function. by using game theory, agent’s behavior in a social network can be modeled too. Establishing an explicit logical connection between these two models is not possible. The purpose of this study is to explore effective implicit relationships between these two models through artificial intelligence techniques.
    Designing a general framework is required for investigation of this correlation and optimizing objective function in social networks with respect to game theory models as behavior of agents. this relation is achieved through free parameters, which are defined in game theory, and searching for the optimal value with respect to objective function. The optimized parameters pattern could be learned by machine learning techniques.Features which are required in learning process are extracted from structural indexes in social network. The learned pattern could be utilized in other social networks to achieve optimal parameters for maximizing objective function
  9. Keywords:
  10. Game Theory ; Multiagent System ; Social Networks ; Increasing Efficiency ; Anarchy Price ; Action Modeling ; Learning Multi Agent Systems

 Digital Object List

 Bookmark

...see more