Research On Uncertain Linguistic Multiple Attribute Group Decision Making Method And Its Application

In practical decision-making problems, such as project evaluation, per-sonnel assessment, scientific research evaluation, supplier selection and enter-prise performance evaluation, due to the influence of subjective factors, such as knowledge structure, personal preferences and working background, etc., and the influence of complexity, fuzziness and uncertainty of the thing itself, experts tend to express their individual preference in uncertain linguistic infor-mation, such as [good, very good]. Studies of uncertain linguistic preference information has caused extensive attention and great interest of scholars both at home and abroad. Introducing the uncertain linguistic information into the multiple attribute group decision making problems can better reflect and em-body the human thinking mode and fully use the wisdom of the expert group, so it is more suitable for solving some complex decision problems in modern society. Research on multiple attribute group decision making method based on uncertain linguistic information has important theoretical significance and high practical value. On the basis of existing research, for the multiple at-tribute group decision making under the environment of uncertain linguistic information, intuitionistic uncertain linguistic information and interval-valued intuitionistic uncertain linguistic information respectively, by using informa-tion entropy, TOPSIS method, nonlinear programming method and rough set theory this paper conducts a series of research in group consensus analysis, the objective comprehensive weighting methods of attributes and experts, the stage weighting model and the evaluation of group decision results, etc., the main research results are as follows:(1) For the consensus problem of uncertain linguistic multiple attribute group decision making, an interactive group decision making method with adaptive experts’weights and feedback mechanism is proposed. In this paper, dominance index is defined to simplify the comparison and ranking process of uncertain linguistic variables, closeness is defined to measure the similar de- gree of two uncertain linguistic variables, based on closeness the multi-level consensus indices are established. Furthermore, according to current consen-sus indices of each expert at all levels, an optimization model is developed to adjust the experts’weights adaptively, and the values that need to be adjust-ed and their reference values are feedback to the corresponding experts which provide targeted guidance for the experts to adjust their individual preference. After multiple iterations, the modified experts’opinions are integrated. Finally, based on the dominance indices all of the alternatives are ranked and the opti-mal alternative can be obtained. The above method can effectively accelerate the consensus reaching process and make the decision results more reasonable.(2) For the multiple attribute group decision making problem in which the weights of attributes and experts are completely unknown under the intuition-istic uncertain linguistic environment, the multi-stage group decision making method and the objective comprehensive weighting model are presented re-spectively. By considering the dynamic characteristics of the decision making process, the former method introduces the logarithmic growth model to de-termine the weights of stages, uses the subjective and objective comprehensive weighting method to determine the weights of attributes and extends the TOP-SIS method to determine the weights of experts. Then by using the aggregation operator and the expectation function, the comprehensive ranking results of al-ternatives are obtained. This method can be used to solve the multiple attribute group decision making problems under the dynamic and fuzzy environment. The latter method determine the objective comprehensive weights of attributes and experts based on the comprehensive consideration of both size and dis-tribution of the values within the single expert and between the experts, the weights information can reflect the relationship of the opinions between the experts more comprehensively and objectively and describe the importance of the attributes and experts better.(3) For the multiple attribute group decision making problem in which the experts’weights are completely unknown under the interval-valued intuition-istic uncertain linguistic environment, a multi-stage group decision making method based on stage weights optimization model and experts’fine-grained objective comprehensive weighting method are presented respectively. The former method uses uncertainty to compute the experts’weights, establishes nonlinear programming model to determine the stage weights and extends the VIKOR method to obtain the ranking results and the optimal alternative. The method provides a new way for solving the similar multi-stage group decision making problems. The latter method thinks the evaluation value of each al-ternative with respect to each attribute under each expert as basic granularity, by defining uncertainty, closeness and rank correlation to describe and measure the experts’weights, the integrated experts’comprehensive weights have high-er accuracy and more comprehensiveness which make the setting of experts’ weights more refinement and rationalization.(4) For the linguistic multiple attribute group decision making problem, the evaluation indices and evaluation criterion for group decision results are proposed. By using the dominance-based rough sets theory, TOPSIS thought and entropy method, the individual decision results and group decision results is analyzed and measured from the structure of ranking, the deviation of value and the distribution of deviation. Thus, the consistency, closeness and unifor-mity are defined and the corresponding evaluation criterion are presented. The method can not only evaluate and rank the expert’s individual decision quality but also reflect the decision effect of group from a certain extent which provides the theory basis for analyzing rationality of the group decision results.(5) On the basis of the existing research of supplier evaluation and selec-tion, the evaluation indices system of materials supplier for coal enterprises is established, the evaluation and selection process of materials suppliers for large coal enterprises is analyzed. Then by using the above uncertain linguistic multiple attribute group decision making methods comprehensively, the mech-anism of ranking and selection of materials suppliers under dynamic and fuzzy environment is developed. Through the analysis of the actual case, the appli-cation value of the methods proposed in this paper is illustrated.The research results of this paper provide a new perspective and solu- tion to the multiple attribute group decision making problems under uncertain linguistic environment, which enrich and develop the theory and approach of linguistic multiple attribute group decision making.