Research On Behavioral Preferences-Based Linguistic Multi-Attribute Group Decision-Making Methods

With the rapid development of modern society,the decision-making environment faced by human beings has become more and more complex.Group decision-making is more appropriate and efficient since it is frequently impossible for a single decision-maker to handle complex decision-making problems independently according to cognitive level and decision-making power limitations.In light of the subjectivity and ambiguity of human thought,multiple attribute group decision making(MAGDM)decision makers are more accustomed to employing qualitative linguistics to convey their opinions and assess data.In addition,decision-makers’ judgments of benefits and losses differ significantly based on their individual cognitive capacity,experience,personality,and preferences.In view this,the study performs a number of novel studies on group decision-making techniques in linguistic contexts,taking into account the behavioral preferences of decision-makers,and the main contents and results are as follows:(1)This paper proposes a multi granularity probabilistic linguistic MULTIMOORA group decision making method based on prospect theory.Firstly,the conversion algorithm of the multi-granular probabilistic linguistic cloud model is given,based on which the basic algorithm,deviation measure,cross-entropy,and fusion tool of multi-granular probabilistic linguistic cloud(MPLC)are proposed.Secondly,an objective model for determining attribute weights and expert weights is established based on the deviation metric and cross-entropy.Considering the ambiguity of ordinal numbers,the final ranking determination method in the MULTIMOORA method is improved to effectively overcome the limitations of the dominance theory.In addition,for the imperfect rationality of decision makers,a multi-granularity probabilistic linguistic cloud MULTIMOORA decision-making method based on prospect theory is proposed.Finally,the applicability and effectiveness of the proposed method are verified by taking wetland ecosystem service value evaluation as an example.(2)In this paper,the bi-direction prospect PROMETHEE II group decision making method under the intuitionistic multiplicative linguistic information environment is proposed.The intuitionistic multiplicative linguistic set(IMLS)is a powerful tool for portraying both qualitative indicators and unbalanced preference performance characteristics.In this paper,we first give the novel score function and the exact function of the intuitionistic multiplicative linguistic variable(IMLV)to obtain the preference relationship of intuitionistic multiplicative linguistic term sets.Meanwhile,the distance measure formula of IMLS is defined,and based on this,the IMLS-CRITIC attribute weight determination model is proposed.Second,given the bi-directional characteristics of membership and non-membership in the set of intuitionistic multiplicative linguistic terms,there are obvious shortcomings in using a single incremental or decremented transformation of the classical Choquet integral operator to gather the intuitionistic multiplicative linguistic information.For this reason,the intuitionistic multiplicative linguistic bi-direction exponent Choquet integral(IMBECI)operator is proposed in this paper.Finally,an improved preference function for the PROMETHEE II method is proposed based on the prospect theory,and the bi-direction prospect PROMETHEE II method under the intuitionistic multiplicative linguistic information environment is thus established.The scientificity and effectiveness of the proposed decision-making method are further illustrated by a numerical calculation example of a water pollution treatment project performance evaluation.(3)This paper proposes a multi-granularity probabilistic linguistic cobweb area-MARCOS(CA-MARCOS)group decision making method based on regret theory.Firstly,this paper gives the conversion algorithm of the multi-granularity probabilistic linguistic term set with normal fuzzy numbers and proves several good properties that it satisfies.The score function,exact function,and distance measure of the normal fuzzy number are defined,and the normal fuzzy number Dombi weighted averaging(NFDWA)operator is proposed.Secondly,considering the intrinsic correlation among experts,the relative importance of each expert is measured by using the proportional Shapley value,and an improved ITARA weight determination method is established based on attribute conflict.In addition,the shortcomings of the existing MARCOS method are effectively overcome by the spider web model,and a highly adaptive CA-MARCOS method is thus proposed.A multi-granular probabilistic linguistic CA-MARCOS method based on regret theory is constructed by considering the risk preference types of experts comprehensively.Finally,the effectiveness and rationality of the proposed method are verified by taking the evaluation of the high-quality development of digital villages as an example.