An Extended Kano Model-and Large-Scale Group Decision Making-Based QFD Method And Application

Quality Function Deployment(QFD)has been proved to be an effective method to integrate customer requirements(CR)into technical characteristics(TC)improvement of products/services.Its core House of Quality(HOQ)can transform the final importance of CRs into the final importance of TCs through a series of relationship matrices,thus guiding the design of new products/services.However,in the face of the increasingly competitive market,the effectiveness and efficiency of the traditional QFD method in practical application are greatly reduced.Take the e-commerce market as an example,on the one hand,the CRs identification and weight determination methods of traditional QFD are increasingly mismatched with the e-commerce economy,which makes it impossible to use the effective design information contained in online shopping reviews.At the same time,when it is used with Kano model,it is impossible to quantitatively determine the CR category,and can not get rid of the dependence on Kano questionnaire;On the other hand,product design based on traditional QFD is more and more unsuitable for the organizational structure of modern enterprises.The network corporate structure of modern enterprises enables more and more experts to participate in decision-making.The priority ranking of TCs is increasingly tending to large scale Group Decision Making(LGDM).Based on this,this paper attempts to propose a new QFD method.It mainly includes the following contents:(1)Extend Kano model with big data technology to quantitatively determine the category and weight of CRs,which solves the problem that traditional QFD methods cannot effectively use the design information in online shopping reviews.First of all,in the CRs identification stage,we use the web crawler technology to collect the design information in online shopping comments,get the customer’s key needs,and determine the objective weight of each CR according to the word frequency of each key demand;Secondly,in the classification stage of CRs,based on the idea of TF-IDF(Term Frequency-Inverse Document Frequency)algorithm,Kano model is expanded by using text analysis,emotion analysis and other technologies,so that it can quantitatively determine the Kano category of each CR according to statistical data,and then determine the standardized weight of CRs;Finally,in the weight adjustment stage of CRs,dynamically adjust the objective weight of CRs according to the enterprise life cycle theory,the Kano category of CRs and the development stage of the enterprise,and finally formulate the standardized weight of CRs suitable for the current development of the enterprise.(2)A large group decision-making method based on clustering and maximum consensus is proposed to help enterprises reach consensus in the TCs ranking stage,and solve the problem that traditional QFD methods cannot adapt to the modern enterprise network organizational structure.Firstly,in the cluster stage of large group decision makers,using social network analysis technology,a network division method based on the social trust relationship of decision makers is proposed,which divides large group decision makers into several subgroups without intersection,reducing the dimension of decision-making problems;Secondly,in the sub group weight determination stage,multiple factors affecting the sub group weight are comprehensively considered.The consistency index and similarity index of the sub group are obtained based on preference analysis,the centrality index of the sub group is obtained based on social network analysis,and the comprehensive weight with parameters of the sub group is obtained by linear combination of them;Finally,a parameter determination method based on the maximum consensus degree is proposed to obtain the comprehensive weight of sub groups,the consensus level among sub groups and the final scheme ranking.(3)The above two methods are organically combined with QFD,and take an ecommerce enterprise as an example to prove its effectiveness.First of all,in the weight determination stage of CRs,it is combined with the data-driven Kano model to quantitatively determine the Kano category and standardized weight of each CR according to the online shopping theory,so as to get rid of the dependence on traditional questionnaires;Secondly,in the TCs ranking stage,combined with the clustering-and maximum consensus-based large-group decision-making method,the weight of each TC and the consensus among experts are determined through the proposed large group decision-making model based on clustering and maximum consensus,which is consistent with the current situation that product design often involves experts in many fields;Finally,combined with the reality,a complete product design process is given using the QFD method proposed in this paper to prove the feasibility and effectiveness of the proposed method.Through the above contents,this paper hopes to enrich the methods and theories of combining QFD with other models,and provide some new ideas for the application of QFD in the current new e-commerce business model.