Research On Supplier Decision-making Problem Driven By Big Data

With the development of the information age,big data has entered various fields,matching and connecting different enterprises in different regions,realizing the rapid transmission and timely feedback of data between enterprises.In addition,the globalization of manufacturing factory is discretization,the production process will involve many different parts of the equipment and many more suppliers,thus the production efficiency and quality has brought the huge impact.Therefore,how to choose the proper one of suppliers is the key of enterprise development.To this,on the basis of predecessors' research results this paper summarized the current supplier decision exits the following problems: the type of evaluation value is single so that it can not accurately reflect the actual circumstances of the various index,and the data is mainly determined by expert scoring method subjectively causing a strong subjectivity result.In view of the problem,the author established a comprehensive index system of supplier under the big data systems,adopted the combination of theoretical research and case verification,in-depth studied VIKOR method of different value attribute to established a decision-making model for suppliers and designed a supplier management system.The following are the main findings and conclusions of this paper:(1)the evaluation values of the five types,such as real number,interval number,triangular fuzzy number,trapezoidal fuzzy number and linguistic variable are studied,and the commensurability between different evaluation values was solved.The linguistic variable is transformed into trapezoidal fuzzy number,and the evaluation values of all attributes are normalized through vector transformation method to eliminate the influence of dimensionality.(2)the aggregation method of subjective evaluation value was studied and the aggregation of subjective evaluation values among all experts was realized.In determining the weight of experts,testing the trust degree of experts,and selecting the most reliable evaluation matrix to determine the weight of experts by maximizing the deviation firstly.Secondly,the weight of the expert is used to assemble the triangular fuzzy number and trapezoidal fuzzy number,which can ensure the objectivity of expert evaluation.(3)a multi-attribute decision-making model based on the VIKOR method was established for the selection of multiple suppliers.Firstly,the weight of all evaluation indexes were calculated by entropy method.Secondly,the "group benefit value" S of the supplier,"individual regret value" R and the interest ratio value Q are calculated by using the method.Finally,the optimal solution is selected based on S,R and Q.(4)the supplier management system was designed.Using JSP+Servlet+Java Bean,Mysql5.0 database and Tom Cat7.0 server,a B /S structure supplier selection evaluation website was built.The site consisting of the enterprise users,suppliers,experts and administrators,mainly has the following functions: enterprise users can choose according to their own needs to choose suppliers according to the recommend sorting system,independent distribution of index weight or the word-of-mouth,three supplier sorting way;supplier upload their information;Experts and enterprise users can make subjective evaluations of suppliers;the administrator can manage enterprise users,suppliers,experts and other information.This paper,by comparing with TOPSIS algorithm,validates the effectiveness of the decision-making model by taking the supplier of auto parts as an example.