The Study Of Group Aggregation Method Based On Linguistic Evaluation Value

Decision-making is an activity of rational people. As the development of social and economy, there are more and more complex decision-making problems. So many of them can't be done with by a single person and they must be done through the wisdom of crowds. Decision-makers often express their preference with linguistic information for ambiguity and uncertainty. So the group-decision making research of linguistic information has attracted the attention of relevant scholars. The conversion mode of linguistic information is proposed in this paper in views of linguistic information. Base on this, the research of group-decision making of single person, multiplayer, large-scale decision makers and multi-stage are studied. The main results are showed as follows.(1)The conversion mode of three-point interval number is proposed. The conversion method of linguistic information and the three-point interval number, the attribute weight model of the three-point interval number and the TOPSIS method are proposed in this paper in views of the interval weights.(2)The alternatives aggregated approach of two-tuple linguistic information process is proposed in views of the linguistic evaluated information group decision-making. Firstly, the linguistic evaluated information is transformed to two-tuple linguistic comparison matrix. Then according to the definitions of two-tuple linguistic and the preferences of policy-makers'comparison matrix, an approach of calculating the policy-makers'attribute weights is proposed using the linear programming model. Secondly, the policy-makers'weights can be obtained through aggregating and comparing the preference of every policy-maker and the whole preference. Finally, the alternative raking can be aggregated.(3)In view of large-scale group aggregation of linguistic assessment information, a new group aggregation method based on two-type linguistic is given in this paper. Firstly, the linguistic information of decision-makers has to be transformed to comparison matrix of two-type linguistic. Secondly, the large-scale decision-makers have to be clustered based on the maximum tree cluster of grey relation. Then the programming models of intra-class attribute weights, the weights of decision-makers and the weights of inter-class are set up. Finally, the consistency of decision-makers is studied.(4)The decision making of multi-stage linguistic information is studied in this paper. In views of uncertain weights of stages, alternatives and decision-makers, the models of dealing with these weights are established. Through these weights, the order of these alternatives can be aggregated.