Simulation Of Dynamic Group Decision Based On Knowledge And Opinion Dynamic

Research on the theories and methods of group decision making (GDM), especially of the dynamic group decision making, is incomplete. There are still several problems need to be solved, e.g. the negligence of knowledge's function in decision making, the lack of systematic and clear description of the process of interaction, the vague conception of the mechanism and the lack of methods in improving decision making performance. Therefore, the focus of this paper is to look into the mechanism of knowledge's function in decision making and build the model of dynamic group decision making based on knowledge and opinion dynamics.Categorizing the GDM according to the convergence mode, our research try to know how the decision making performance is affected by knowledge.Firstly, the definition of opinion dynamics and its classic models are introduced and the dynamic of group decision making is descriped by group opinion evolution. The definition of knowledge and its configuration is described in the form of weighted network, knowledge potential is introduced to explain how knowledge transfer happens and its definition.Secondly, the continuous inputting of knowledge has a positive impact on the precision of description on the decision problem. A new mechanism is proposed: knowledges are rules to distinguish and decide, the more knowledge and its potential, the more the rules to be used and the higher precision of decision, the ratio of knowledge quantities in different areas affects the decisioner's preference, knowledge transfer continually activate the decision rule set so that the performance is improved.Of all the factors affecting decion making, the knowledge potential difference is determined to be an important factor in knowledge transfer, then the conceptual model of dynamic group decision making is proposed in the basis of knowledge and opinion dynamic. The conception of dementional knowledge ratio and knowledge weight is introduced, together with the method to calculate the knowledge potential, to build the model of dynamic GDM in three different stages which form an integrated system. The model is carried out in Matlab and runed in the initial setting. Finally, the influence of group qualities, eg, the number of people, the distribution of initial knowledge and the knowledge weight is simulated and analyzed; the affect of different mechanism of knowledge transfer is analyzed because the new model is based on knowledge. Knowledge creation is took into consideration in the model and simulated under different initial knowledge distribution.