Loading...

Multi Agent Systems for Modeling the Opinion Formation Phenomenon in Complex Dynamical Networks

Askari Sichani, Omid | 2013

801 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 45676 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Jalili, Mahdi
  7. Abstract:
  8. Opinion formation is one of the major topics in social networks analysis and mining. A number of methods have been introduced for describing and simulating this phenomenon. In this thesis, we proposed a novel optimization framework based on an opinion formation model that used the Deffuant’s model with some changes in neighbors’ selection policy. Its efficiency was compared with a number of benchmark optimization methods including genetic algorithms, differential evolution and particle swarm optimization.The proposed model demonstrated better performance than the others.In the proposed, each person changes his opinion based on one of his neighbors who has the best performance in terms of the desired cost function. We proved that if some conditions are met, convergence of the optimization model is guaranteed.As another application of opinion formation models, we proposed a framework for uncovering the influential individuals as well as their social power in a networked structure. The problem was converted to a Bayesian probability estimation problem and represented as a concave function which can be easily maximized. We apply the proposed method on a number of model and real networks and show that the estimated values of the social power, are highly correlated with the real values. We also consider an application of finding influential nodes in opinion formation through informed agents. In this application a number of informed agents are attached to people who have the most social power and try to change society’s will to their desired. Our numerical simulations show that the proposed method outperforms many other heuristic methods. We also used the proposed framework for another type of information diffusion modeling based on Hawkes process’ methods. For further investigation about finding the best spreadernodes in a more realistic environment, we used an extended version of the bounded confidence model in which the uncertainty of each agent is adaptivelymight be changed by the network. As before, informed agents create links with the residents to promote their opinions. As a result, we proposed to connect them to the nodes with small-degrees that are connected to high-degree nodes. Our experimentals on multiple networks show the superior performance of the proposed metric over the state of the art heuristic methods
  9. Keywords:
  10. Information Diffusion ; Optimization ; Convex Optimization ; Bayesian Learning ; Multiagent System ; Community Detection ; Maximum Likelihood Estimation ; Complex Dynamical Networks ; Opinion Formation ; Stochastic Matrix

 Digital Object List

 Bookmark

No TOC